A High Binary Fraction for the Most Massive Close-In Giant Planets and Brown Dwarf Desert Members

Home , Altair, Arcturus, HD 104985, HD 104985 b, HD 114762, HD 114762 b

MNRAS 000,1–30 (2019) Preprint 29 March 2019 Compiled using MNRAS LATEX style ﬁle v3.0

A high binary fraction for the most massive close-in giant planets and brown dwarf desert members

C. Fontanive1,2?, K. Rice1,2, M. Bonavita1,2, E. Lopez3,4, K. Mužic´5 and B. Biller1,2

1SUPA, Institute for Astronomy, University of Edinburgh, Blackford Hill, Edinburgh EH9 3HJ, UK 2Centre for Exoplanet Science, University of Edinburgh, Edinburgh EH9 3HJ, UK 3NASA Goddard Space Flight Center, 8800 Greenbelt Rd, Greenbelt, MD 20771, USA 4GSFC Sellers Exoplanet Environments Collaboration, NASA GSFC, Greenbelt, MD 20771, USA 5CENTRA, Faculdade de Ciências, Universidade de Lisboa, Ed. C8, Campo Grande, P-1749-016 Lisboa, Portugal

Accepted 2019 March 5. Received 2019 February 27; in original form 2019 January 17

ABSTRACT Stellar multiplicity is believed to influence planetary formation and evolution, although the precise nature and extent of this role remain ambiguous. We present a study aimed at testing the role of stellar multiplicity in the formation and/or evolution of the most massive, close- in planetary and substellar companions. Using past and new direct imaging observations, as well as the Gaia DR2 catalogue, we searched for wide binary companions to 38 stars hosting massive giant planets or brown dwarfs (M > 7 MJup) on orbits shorter than ∼1 AU. We report the discovery of a new component in the WASP-14 system, and present an independent con- firmation of a comoving companion to WASP-18. From a robust Bayesian statistical analysis, +13.2 we derived a binary fraction of 79.0−14.7% between 20−10,000 AU for our sample, twice as high as for field stars with a 3-σ significance. This binary frequency was found to be larger than for lower-mass planets on similar orbits, and we observed a marginally higher binary rate for inner companions with periods shorter than 10 days. These results demonstrate that stellar companions greatly influence the formation and/or evolution of these systems, suggesting that the role played by binary companions becomes more important for higher-mass planets, and that this trend may be enhanced for systems with tighter orbits. Our analysis also revealed a peak in binary separation at 250 AU, highlighting a shortfall of close binaries among our sample. This indicates that the mechanisms affecting planet and brown dwarf formation or evolution in binaries must operate from wide separations, although we found that the Kozai-Lidov mechanism is unlikely to be the dominant underlying process. We conclude that binarity plays a crucial role in the existence of very massive short-period giant planets and brown dwarf desert inhabitants, which are almost exclusively observed in multiple systems. Key words: planetary systems – planets and satellites: formation – binaries: visual – binaries: close – methods: observational – methods: statistical

1 INTRODUCTION the characteristics and demographics of planetary populations (e.g.

arXiv:1903.02332v2 [astro-ph.EP] 28 Mar 2019 Desidera & Barbieri 2007; Eggenberger et al. 2007, 2011; Daem- In the search for analogues to the planets in our own Solar System, gen et al. 2009; Adams et al. 2012, 2013; Ginski et al. 2012). The exoplanet studies originally firmly excluded known binary systems, dominant results that emerged from these surveys were a strong despite the fact that about half of Solar-type stars are found in mul- deficit of binary companions within ∼50−100 AU for planet hosts tiple systems (Raghavan et al. 2010). Serendipitous discoveries and (Roell et al. 2012; Bergfors et al. 2013; Wang et al. 2014a,b; Kraus subsequent dedicated surveys later revealed that a significant frac- et al. 2016), and the idea that massive short-period planets appear to tion of exoplanets are actually found in binary star systems (e.g. be preferentially found in multiple-star systems (Zucker & Mazeh Patience et al. 2002; Desidera et al. 2004; Mugrauer et al. 2006; 2002; Eggenberger et al. 2004). Mugrauer & Neuhäuser 2009), mostly with binary separations of at least a few hundred AU. These findings led to numerous stud- These studies, however, focused primarily on systems in ies investigating how stellar binarity affects planet formation and which the planet had a mass less than ∼4 MJup. Theoretical calcula- tions (Kratter et al. 2010; Forgan & Rice 2011) and numerical simulations (Stamatellos & Whitworth 2008; Stamatellos 2013; Hall ? E-mail: [email protected] et al. 2017) both suggest that planets that form via disc fragmen-

c 2019 The Authors 2 C. Fontanive et al. tation in gravitationally unstable discs (Boss 1998) typically have jects within the brown dwarf mass regime, have yet to be studied in masses above ∼4 MJup. Therefore the planets in these existing stud- detail in the context of stellar multiplicity. Zucker & Mazeh(2002) ies probably formed via the standard core accretion scenario (CA; were the first to point out that massive (Mp > 4 MJup) short-period Pollack et al. 1996), rather than via gravitational instability (GI). planets tend to be predominantly found orbiting one component of When it comes to planets that formed via core accretion, bina- a multiple star system and possess distinctive characteristics com- rity on close separations is generally considered to have a negative pared to planets orbiting single stars (Eggenberger et al. 2004). influence (see Thebault & Haghighipour 2015 for a review of planet Such massive planetary and substellar companions are very formation in binaries and the issues introduced by the presence of challenging to form at small separations. Giant planet formation, a close binary companion). Theoretical studies have concluded that whether by CA or GI, is thought to occur preferentially in the rel- close stellar companions can hinder planet formation by stirring up atively cool outer regions of protoplanetary discs, from a few AU protoplanetary discs (e.g. Mayer et al. 2005), tidally truncating the for CA (Pollack et al. 1996), to several tens of AU for GI (Rafikov discs (e.g. Pichardo et al. 2005; Kraus et al. 2012), or leading to the 2005). Massive hot Jupiters are thus expected to have formed at ejection of planets (Kaib et al. 2013; Zuckerman 2014). More com- wide orbital separations from their host stars, where the lower tem- pact, truncated discs generally have just enough mass left to form a peratures in the protoplanetary disc allow for planet formation to low-mass Jovian planet (Jang-Condell et al. 2008), and planet for- proceed (Bell et al. 1997), or be born under very different condi- mation is then further complicated by the very short lifetime (.1 tions than currently encompassed by most planet formation mod- Myr) of these truncated discs (Kraus et al. 2012). els. Recently, Schlaufman(2018) found evidence for two distinct On the other hand, Batygin et al.(2011) and Rafikov(2013) populations of close-in giant planets, with a suggestion of a tran- predict that stellar companions should have little influence on plan- sition between CA and GI companions occurring at around ∼4−10 etesimal growth. It has also been proposed that the presence of an MJup. This is consistent with both semi-analytic (Kratter et al. 2010; outer companion could raise spiral arms in protoplanetary discs, Forgan & Rice 2011) and numerical simulations (Stamatellos & creating regions of high gas and particle densities, favourable for Whitworth 2008; Stamatellos 2013; Hall et al. 2017) which sug- planetesimal formation (Youdin & Goodman 2005) and pebble ac- gest that objects that form via GI have masses above ∼3−5 MJup, cretion (Johansen et al. 2007; Lambrechts & Johansen 2014). For and might suggest that these more massive close-in planets formed example, the spiral arm structures observed in the disc around HD by GI rather than via CA. 100453 (Wagner et al. 2015) may be due to perturbations from the In this work, we aim to constrain the multiplicity statistics of M-dwarf companion (Dong et al. 2016), located at 120 AU from hosts to the most massive giant planets (Mp > 7 MJup) and brown the primary and originally reported by Chen et al.(2006). Simi- dwarfs found within ∼1 AU, in order to investigate the role of bina- larly, the asymmetric disc of HD 141569 is attributed to the stellar rity in the formation or evolution of these systems. This will allow companions in this triple system (Augereau & Papaloizou 2004). us to assess if a wide binary companion could be responsible for In the “Friends of hot Jupiters” series of papers, Knutson et al. the observed orbital configurations of these objects, which are both (2014), Piskorz et al.(2015) and Ngo et al.(2015, 2016) found scarce and challenging to explain with current formation theories. a binary fraction 3 times higher for hosts to hot Jupiters (mostly Our investigation will provide an indication of whether the Kozai- up to Mp ∼ 4 MJup) than for field stars on separations of 50−2000 Lidov mechanism could play a role in the origin of the most mas- AU, and concluded that wide binary systems may either facilitate sive short-period gas giant planets and brown dwarfs. This study the formation of Jovian planets, or help the inward migration of will also help us determine if these massive companions are an planets formed at wider separations. extension of the population of lower-mass, CA giant planets, or It has also been suggested that binary companions could in- if they belong to a separate population formed through a distinct duce the inward migration of planets through secular interactions mechanism (i.e. GI on wide orbits, followed by inward migration; such as the Kozai-Lidov mechanism (Kozai 1962; Lidov 1962). In Nayakshin 2010; Rice et al. 2015). In particular, we will explore the this scenario, an outer companion with a large mutual inclination binary properties of hosts to members of the “brown dwarf desert” between the planetary and binary orbits can excite large periodic (Marcy & Butler 2000), depicting the significant deficit of brown oscillations of the eccentricity and inclination of the planet. Tidal dwarf companions found within a few AU around Sun-like stars interactions between the planet and its host star can then drive the (e.g. Grether & Lineweaver 2006; Ma & Ge 2014). planet onto a final orbit with a very small orbital separation when In Section2 we present our selected sample of targets. Sec- compared to its initial location (e.g. Fabrycky & Tremaine 2007; tion3 describes the direct imaging observations acquired for this Naoz et al. 2012; Petrovich 2015). In particular, the Kozai-Lidov project and the data reduction. In Section4 we detail our search migration process has been proposed to explain the high obliqui- for wide companions, using past imaging surveys found in the lit- ties often observed in hot Jupiters, although recent studies indicate erature to complement our new direct imaging data, as well as the that this mechanism can only account for about 20−30% of the ob- Gaia Data Release 2 (GR2; Gaia Collaboration et al. 2016, 2018) served hot Jupiter population (Naoz et al. 2012; Ngo et al. 2016). catalogue. Section6 describes our approach to the statistical anal- Similarly, it has been suggested (Rice et al. 2015) that Kozai-Lidov ysis of our survey, and we present our results in Section7. Finally, oscillations could drive planetary-mass bodies that form on wide we discuss our interpretation of the obtained results in Section8 orbits via GI onto short-period orbits. Since disc fragmentation and summarise the main results of our project in Section9. preferentially forms massive planets or brown dwarfs (Kratter et al. 2010; Stamatellos & Whitworth 2008; Forgan & Rice 2011), such a process would tend to be associated with more massive planets 2 SAMPLE SELECTION (Mp & 4 MJup), or brown dwarfs. Although the true influence of binarity on planet formation The aim of this project is to search for wide, substellar or stellar and evolution is still unclear, systems hosting planets with masses companions to stars hosting a massive planet or brown dwarf on a up to a few Jupiter masses have been extensively surveyed. In con- very short orbit. Recent findings suggest that GI forms planets with trast, systems with more massive planets (Mp > 4 MJup), and ob- masses larger than ∼4 MJup (Stamatellos 2013; Stamatellos & Her-

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 3 Ref. [yr]) circ 10 8.0 Siverd et al. ( 2012 ) τ 0035006 14.5 06Butler 15.9 et al. ( 2006 ) 11.9 McArthur et al. ( 2010 ) Almeida et al. ( 2017 ) 09201009 16.7 12.5 Guenther et al. ( 2009 ) 15.6 00043Galland et al. ( 2006 ) 160019Mitchell 14.9 et al. ( 2013 ) 10.2 Wilson et al. ( 2016 ) 9.4 Johns-Krull et al. ( 2016 ) Nowak et al. ( 2017 ) 005 15.9 Liu et al. ( 2008 ) 024 16.4 Döllinger et al. ( 2007 ) 01003 15.8 15.5 Sato et al. ( 2010 ) Jones et al. ( 2014 ) 00015 8.0 Esteves et al. ( 2015 ) 0028017 8.0 8.4 Hellier et al. ( 2009 ) Winn et al. ( 2008 ) 002001003 10.2 12.6 14.4 Udry et al. ( 2002 ) Udry et al. ( 2002 ); Wittenmyer et al. ( 2007 ) Wittenmyer et al. ( 2009 ) 0030 8.1 Wong et al. ( 2015 ) 000500009 13.1 14.5 Kane et al. ( 2011a ) Wilson et al. ( 2016 ) 002018 11.2 12.9 Wilson et al. ( 2016 ) Díaz et al. ( 2012 ) 0048 14.2 Kane et al. ( 2011b ) 02202 16.4 16.8 Johnson et al. ( 2011 ) Galland et al. ( 2005 ) 0033005 8.5 008 8.3 012Pál et al. ( 2010 ) 00036Bakos et al. ( 2011 ) 16.9 02007 8.0 13.3 009Wilson et( 2016 ) al. 11.4 Wilson etZucker al. ( 2016 ) et al. ( 2004 ) 14.7 16.8 Ma et al. ( 2016 ) Wittenmyer et al. ( 2009 ) Sato et al. ( 2008 ) 0100 0069 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 02 8.8 Irwin et al. ( 2010 ) . + − 23 16.1 Jones et al. ( 2013 ) ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± . 0 < 34 34 20 28 05 02 ...... 0099 245 289 031 231 432 257 260 277 530 114 891 708 066 015 394 081 673 090 ...... 4007 0092 0830 5967 3354 1929 5171 0 . 0 ...... 00064 84785 50048 69397 . . . . 96 0 5 0 1 0 29 0 09 0 8 0 7 0 22 0 94 0 . 6 0 2 0 . 2728 0 0 . 66 0 61 0 05. 0 4 0 . 53 0 . . 23 0 . i e 6 4 ...... ) (log . . 1 2 3 0 0 1 0 3 0 0 20 0 0 0 2 0 0 0 1 0 0 0 0 0 1 0 2 3 6 0 + − 1 0 . 3 0 + − sin ± ± ± ± ± ± ± ± ± ± ± Jup . . 18 0 ± ± ± ± ± ± ± ± ± ± ± 3 ... 0 . ... . 2 2 4 4 4 . 1 2 . 6 . . 25 . 37 . . 57 98 M 88 48 03 71 . . 49 96 08 44 ...... 1 1397 8 ... 0 6959 ...... 0 0 3 3 . . 24187 ...... 0 0 16 ... 0 . . 47 ... 0 . . 50 48 8 6 ...... ) (M . . 2 5 2 0 0 0 0 0 2 2 0 0 0 0 + − 2 + − + − ± ± ± ± Jup ± ± ± ± 7 ... 20 98 . 23 . . 29 84 30 79 09 . . 28 76 . . . . . 246 . 04 ... 19 . 03 ... 13 012028 ...... 08 9 00068 7 0005 ... 9 02 7 53 013 ...003010 ... 7 ...... 41 12 8 0043 13 0029 36 05 ... 19 04 ... 7 0300016 27 ... 8 00087 9 001 33 005006 ...... 0011 7 00068 7 7 00082 10 11 020043 ...... 14 48 017021 ...... 9 10 010 ... 47 001 ... 10 ...... 02 08 . . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 ... 9 + − ± . 17 ... 20 95 ... 8 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± . . 565 ... 9 0177. ... 18 . 34 . 08 50 76 29 87 81 68 . . . . 162 . . . . 484 438 117 918 995 024 290 339 495 096 638 353 ...... 1 . . . . . 0361 8259 0593 0371 0751 . . . . . 02466 03641 02026 04540 06878 . . . . . 047 0 1 1 0044 0 . 5 0 32 1 96 0 5 0 62 1 0074 0 540071 0 1 004 0 016 0 085 0 320026 0 05 0 0 . 0009 0 002 0 . . . . 03. 0 . 0030 0 . . . 001. 0 . 01 0.2 ... 25 0 . . . . 4 2 . 0 000015 0 000001 0 000025 0 . . 000056 0 . . . 000004 0 000012 0 0000061 0 0202000026 0 0.079 12 00000044 0 0000068 0 00000087 0 . . . 0 0 0 0 3 1 0 0 0 . 0 . 0 0 0 . . . 2 0 1 . 0 . . . . 0 . 1 1 0 0 0 0 0 0 6 0 0 0 0 0 0 0 0 0 0 0 0 0 ± + − . ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 3 P a M ± ± ± ± ± ± ± ± ± ± ± ± . 5 1 2 . . . 03 30 03 05 11 78 (days) (AU) (M . . . . 84 . . 44 . . 680 081 505 484 . . . 481 9402 6884 . . 9049 5546 1135 . 1112 . . . 9151 . . . 9891 . 217514 763588 428198 875317 328300 292569 889475 ...... 6334729 1915239 . . 24376524 94145299 . . p N is the number of known planets in the system. Tidal circularisation timescales were calculated in this paper (see text). All other parameters come from the given references and references therein. p N Orbital properties of the planets considered. 24 4697 b 1 145 + And c 4 240 Gem b 1 305 59 Dra b 1 28 30 Ari B b 1 335 11 Com b 1 326 Planet ID 4 UMa b 1 269 Notes. HD 203949 bKELT-1 b 1 1 184 1 HD 168443 bHD 178911 B bHD 180314 b 2 1 1 58 71 396 HD 160508 bHD 162020 b 1 1 178 8 HD 156279 bHD 156846 b 1 1 131 359 HD 134113 b 1 201 HD 33564 bHD 39392 b 1 1 388 394 τ AS 205A bBD 1HAT-P-20 bHD 5891 b 24 1HD 41004 B bHD 77065 b 1HD 2 87646 A bHD 89744 b 1HD 104985 b 177 HD 1 112410 b 2 1 HD 114762 b 1 1 119 1 13 1 256 199 83 124 70 Vir bυ CI Tau 1 bEPIC 219388192 bHAT-P-2 b 1 116 5 1 1 5 8 Kepler-13 A bNLTT 41135 b 1 1 1 2 WASP-14 bXO-3 b 1 2 1 3 WASP-18 b 1 0 Table 1.

MNRAS 000,1–30 (2019) 4 C. Fontanive et al. (15) HIP 25110 (14) HIP 4715 (13) (11) (33) (30,31) 1 (17,18) HIP 49522 66 (2) HIP 74033 6 (6,22) HIP 64426 30 (6,25) HIP 89844 . 34. (21) HIP 63242 4069 (2)30 HIP (2) 27828 2 (6,19) HIP 44259 . HIP (6,20) 50786 HIP 58952 20 (6,23) HIP 84171 200 (4) HD 180777, HIP 94083 37 (2,21) HIP 84856 02883 (1) (2)4 (3) HD 73108, HIP HD 42527 107383; HIP 60202 HD (5,6) 16232, HIP 12184 HD 117176, HIP 65721 57. (2) HIP 86394 76 (7) HD 54719, HIP 34693 5 (12) HD 147506, HIP 80076 70 (6,24)30 HIP 87330 (6,19) HIP 94075 22 (8) HD 9826, HIP 7513 24 (2) HIP 94576 150 (2) HIP 113698 19 (21) HIP 105854 25 (26) 20 (6,32) HD 10069, HIP 7562 10 (27,28) KOI-13 ...... 82 58 5 8 3 6 5 8 8 7 0 5 . . 1 0 0 0 ...... 2 1 3 0 1 2 0 0 2 0 0 0 0 0 0 2 1 0 0 4 0 0 0 0 0 0 0 0 0 3 0 1 5 1 1 1 − + ± ± ± ± − + − − + − + + − + ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 6 4 0 0 7 4 Age Ref. Other name . . . . . 9 . 6 6 1 82 75 98 . . . 00 . . 17 06 59 . 50 . 40 78 17 91 55 22 10 . 80 12 14 23 75 90 12 ...... 436 207 . . 15 4 01 12 25 4 1 4 048 2 02 7 04 9 01 7 09 4 02 10 05 2 . 1104 1 01 0 8 04 3 04 5 015 0 3 1 0102 3 4 041 1 02 3 01 10 13 1 04 2 028 6 10 1 100 0.001 (9) V866 Sco, EPIC 205249328 01602 5 0.002 (10) EPIC 247584113 063 1 04 3 01 0 066 2 10 1 02012 5 2 (29) ...... 0 ) (Gyr) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ∗

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± M 2 3 . . 83 93 94 21 12 85 38 07 14 75 03 31 02 13 36 99 80 01 20 72 35 71 02 16 25 ...... 234 ...... 558 . 447 242 756 086 754 335 213 164 ...... 03 1 030 0 01 0 01 0 05 1 08 1 0202 0 1 02 1 10 2 1103 0 1 08 1 067 1 0507 1 2 08 1 08 0 06 1 060 1 03050 0 0 09 1 080 1 10 1 08 1 04 1 02195 2 1 ...... 02 0 . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 H] 1015 0.4 1 1.6 (16) HIP 28393 12 1 03 1 85 1 . / ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± . . 012 1 016 1 . . . . . 0 ± 0 0 0 0 0 0 0 − − + − + − 26 14 54 28 17 92 18 25 16 31 14 01 34 13 06 24 14 35 28 16 00 50 13 − + ...... 42 774 131 043 727 177 245 . 0 . 0 0 0 0 0 0 0 0 0 . 0 0 0 0 . 0 0 0 0 0 . 0 . 0 . 0 0 0 0 0 0 0 0 + + − − − − + − − − + + + + + + + + + − + − − − + + − − − 6 3 7 7 . 7 8 1 9 5 5 6 2 6 0 4 7 . 8 1 4 1 . . 5 1 0 . 4 1 2 1 . 9 1 1 . 1 1 1 . 1 1 . . 2 8 . 3 ...... 1 ...... 0 0 0 0 4 4 0 0 1 1 4 3 0 2 0 0 0 0 0 0 15 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 18 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 5 5 6 6 5 5 4 8 5 0 0 4 5 1 0 6 2 2 . 7 . 2 6 1 . 7 5 9 . 2 7 0 0 4 . 8 4 . 6 . . 0 6 . 7 . . . . . 3 . 1 ...... Distance [Fe 030 71 21 127 020 305 057 268 037 473 009 38 013 36 013 40 013 102 021 32 01 156 018 72 013 137 009 100 010 47 005 73 009 17 012 111 02011 93 44 009 21 010 41 009 27 022 30 05 13 01 283 011 39 010 122 013. 127 009 78 . 022 44 . . 018 123 027 213 . 026 162 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± V 63 86 74 10 11 . . . . . 346 535 701 349 782 167 361 449 786 339 143 785 564 787 045 177 140 494 020 107 227 000 721 743 . 620 847 . . 357 904 . 798 ...... V 9 / 41:13:46.3 F7 V 5 63:21:07.5 K0 V 8 17:31:01.6 F9 V 7 22:04:19.7 F8 V21:27:13.4 G5 8 V 8 08:52:47.2 F9 V 8 48:14:22.9 M2 V17:53:42.4 G1 IV76:54:20.6 12.6365:38:47.3 G8.5 III 8 G8 III 5 41 6 19:20:01.5 G1 V 6 64:19:40.6 K1 III 4 13:46:43.6 G5 V 5 26:45:27.1 F8 V 8 79:13:52.1 F6 V 5 34:36:01.6 G5 V 7 17:47:34.324:38:53.0 G8 III76:33:37.8 F6 V A9 V 4 30:14:42.6 7 K2 5 III 4.42 112 40:19:06.1 K3 V 9 20:17:33.0 G5 III 8 41:24:19.6 F8 V 4 24:20:11.5 K3 V 11 09:35:44.6 G6 V31:51:37.3 K0 7 III 6 41:02:53.1 F8 V 8 18:38:26.0 K5 V 12 37:49:46.0 K2 III 5 25:47:16.5 K2 V 9 22:50:30.116:52:17.8 K7 V G V 12 13.80 158 39:23:01.8 F5 V 10 45:40:40.4 F6 IV 46:52:05.9 A5 V 10 57:49:01.9 F5 V 9 21:53:41.0 F5 V 9 04:41:30.0 M5.1 V 18 34 + + + + + + − + + − − + + + + + + + + + − + + + − + + − − + + − + − + + + + (J2000) (J2000) (mag) (pc) (dex) (M (1) Döllinger et al. ( 2007 ); (2) Maldonado &( 2017 ); Villaver (3) Guenther et al. ( 2009 ); (4) Jones et al. ( 2015 ); (5) Butler et al. ( 2006 ); (6) Bonfanti et al. ( 2016 ); (7) Mitchell et al. ( 2013 ); (8) Distances are based on the estimates from Bailer-Jones et al. ( 2018 ) derived from Gaia DR2 parallax measurements. All other parameters come from the given references and references therein. Stellar properties for the selected systems. 24 4697 23:01:39.32 + And 01:36:47.84 Gem 07:11:08.37 HD 89744 10:22:10.56 HD 156279 17:12:23.20 HD 114762 13:12:19.74 HD 39392HD 77065 05:53:19.00 09:00:47.45 HD 134113 15:07:46.50 HD 41004 BHD 87646 A 05:59:49.65 HD 104985HD 10:06:40.77 112410 12:05:15.12 12:57:31.96 HD 156846 17:20:34.31 Object ID4 UMa RA70 Vir 08:40:12.82 Dec. 13:28:25.81 SpT HD 160508 17:39:12.70 Notes. References: McArthur et al. ( 2010 ); (9) Almeida et(16) al. ( 2017 );Santos (10) etGuilloteau al. ( 2002 ); et al. ( 2014 ); (17) ( 2012 ); (11) Aguilera-Gómez (24) Nowak et etUdry al. ( 2018 );( 2017 ); al. et (12) (18) al. ( 2002 );PáletMa (25) et al. ( 2014 ); etWittenmyer al. ( 2010 ); (32) et al. ( 2016 ); (13) Hellier al. ( 2007 ); (19) Bakos et (26) etWittenmyer al. ( 2009 ); al. ( 2011 );Siverd et (33) (14) et al. ( 2009 );WinnJohnson al. ( 2012 ); (20) et et (27) al. ( 2008 ). Sato al. ( 2011 );Morton et (15) etGalland al. ( 2008 ); al. ( 2016 ); et (21) (28) ( 2005 ); al. JofréShporer et et al. ( 2015 ); al. ( 2014 ) (22) (29) KaneIrwin et et al. ( 2011b ); al. ( 2010 ); (23) (30) Díaz( 2012 );Southworth et (31) al. Knutson HD 33564 05:22:33.53 HD 178911 B 19:09:04.39 11 Com b30 Ari B59 Dra 12:20:43.03 τ 02:36:57.74 19:09:09.88 HD 162020 17:50:38.36 HD 5891 01:00:33.19 υ HAT-P-20 07:27:39.95 HD 168443HD 180314 18:20:03.93 19:14:50.21 HAT-P-2 16:20:36.36 AS 205A 16:11:31.34 HD 203949 21:26:22.87 BD CI TauEPIC 219388192 19:17:34.03 04:33:52.01 KELT-1 00:01:26.92 WASP-18 01:37:25.03 Kepler-13 A 19:07:53.15 XO-3 04:21:52.71 WASP-14 14:33:06.36 NLTT 41135 15:46:04.26 Table 2.

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 5 czeg 2015; Hall et al. 2017), and the transition between CA and GI for companions discovered through this method only allow for the companions is thought to occur around ∼4−10 MJup (Schlaufman determination of a lower limit on the companion mass due to the 2018). Studies of core accretion populations found that CA rarely unknown inclination i of the system. Radial velocity systems are forms planets with masses larger than ∼5 MJup (Matsuo et al. 2007; therefore expected to be more massive than the estimated M2 sin i Mordasini 2018), and shows a steep drop and a strong metallic- as a result of the projection factor. Selected systems discovered via ity dependence in the formation of higher-mass planets (Mordasini this method are thus likely to be more massive than the minimum et al. 2012; Jenkins et al. 2017). In order to investigate the higher- masses reported in Table1. Given the projected masses of these mass planetary population, which likely formed by disc GI rather companions and assuming a uniform distribution of inclinations be- than through CA, we choose for this survey a lower limit on in- tween 0 and 90 degrees, we can easily show with a Monte-Carlo ner companion mass M2 of 7 MJup, based on the studies mentioned approach that an average of 72% of the radial velocity systems above. This allows us to avoid the main region of overlap between considered here are statistically likely to be above the deuterium CA and GI, while keeping a sufficiently large sample size for our burning limit at 13 MJup. Combining this with our transiting sys- study. We place an upper limit of 70 MJup (the hydrogen-burning tems, this means that more than ∼60% of our targets are likely in limit) on the mass (or projected mass) of the inner companions, so the brown dwarf mass regime, and close to 80% of our sample is as to limit our sample to likely substellar objects. expected to have a true mass >10 MJup. We therefore consider that We place an upper limit of P < 400 days (about 1 AU around a our gathered sample of objects is representative of the population Sun-like star) on the orbital period of the close-in companions. It is of the most massive planets found on tight orbits and provides a now well-accepted that if planet formation via GI does occur, it typ- robust insight into short-period brown dwarf desert members. ically takes place in the outer regions (>30 AU) of protostellar discs We define in Table1 a tidal circularisation timescale τcirc for (Rafikov 2005; Clarke & Lodato 2009; Rice & Armitage 2009). each planet and brown dwarf companion in our sample, given in

This thus ensures that all selected companions have undergone sig- log10[yr]. We estimate this parameter using the formalism pre- nificant migration between their expected GI formation location sented in Rice et al.(2012), which is based on that developed by and their current configurations, or that they had to be formed under Eggleton et al.(1998) (see also Mardling & Lin 2002 and Dobbs- considerably different natal environments than for standard planet Dixon et al. 2004). We assume the star has a tidal quality factor of 6 5 formation in order to be born in-situ. We set an upper limit of 500 Q∗ = 5 × 10 and the planet has a tidal quality factor of Qp = 10 . pc on the distances of our targets in order to be sensitive to wide We take the star mass, planet mass, orbital semi-major axis, and companions from 50−100 AU around most stars in our sample. We orbital eccentricity from Tables1 and2. We assume that the star use the distance estimates from Bailer-Jones et al.(2018) to in- has a rotation period of 20 days and that the planet is rotating syn- fer distances for our targets. These distances are derived from the chronously. We estimate the circularisation timescale by simply highly-precise parallax measurements provided by the Gaia DR2 evolving each system for a short period of time and determining catalogue, correcting for the nonlinearity of the transformation be- the resulting change in eccentricity (i.e., τcirc = e/e˙). tween parallax and distance. Finally, we only consider stellar pri- Petrovich(2015) found that planets migrating via the Kozai- maries and place a limit on the host’s mass of M∗ > 0.1 M . Lidov mechanism (Kozai 1962; Lidov 1962), under the influence of Based on the arguments presented above, we selected all sys- a distant companion, spend most of their lifetimes undergoing ec- tems from the NASA Exoplanet Archive1, the Exoplanet Data Ex- centric oscillations at separations >2 AU, or as Hot Jupiters at <0.1 plorer2 and the Extrasolar Planets Encyclopaedia3 with confirmed AU. All the inner companions considered here have orbital separa- transiting or radial velocity companions with well-constrained or- tions smaller than 1 AU. If they migrated from wider separations bits that satisfy the following criteria: to their current locations through the Kozai-Lidov scenario, they should be able to circularise onto Hot Jupiter orbits fairly rapidly. − − inner companion mass M2 (or M2 sin i) between 7 70 MJup. Inner companions with circularisation timescales longer than the − inner companion orbital period P < 400 days. age of the Universe are thus unlikely to be driven by secular per- − distance within 500 pc based on Gaia DR2 parallax. turbations such as the Kozai-Lidov mechanism. On the other hand, − primary mass M∗ > 0.1 M . objects with timescales smaller than the age of the Universe (i.e., Our final sample consists of 38 objects, and includes some very less than ∼10.2 in log10[yr]) could have migrated inwards via the short period (P < 10 days) transiting systems, together with ra- Kozai-Lidov scenario. A total of 12 targets have tidal circularisa- dial velocity objects extending to larger separations. Properties of tion timescales shorter than that and may thus be consistent with the inner companions are presented in Table1 and the host stars a Kozai-Lidov migration process. The subset of Kozai-consistent are listed in Table2. We selected our sample without regard to the objects corresponds to all the inner companions in our sample with targets’ multiplicity, known or unknown. However, radial velocity an orbital period shorter than 10 days. This is in good agreement and transit surveys are typically biased against binaries, excluding with the idea that planets migrating via the Kozai-Lidov mechanism known multiple systems in target selection processes. As these bi- spend most of their lifetime around their initial, wide separations, ases are difficult to quantify and account for, our obtained results or on hot Jupiter orbits, as discussed in Petrovich(2015). may somewhat underestimate the multiplicity rate of the population probed here, but we consider that our study is in no way biased towards the presence of wide companions. 3 NEW OBSERVATIONS About three quarters of the selected systems were discovered 3.1 Observations and data reduction and characterised via radial velocity measurements. Mass estimates We used direct imaging facilities at the Very Large Telescope (VLT), Gemini North and the WIYN Observatory to acquire data 1 https://exoplanetarchive.ipac.caltech.edu for six objects in the sample presented in Section2, four of which 2 http://exoplanets.org did not have any previously reported direct imaging observations. 3 www.exoplanet.eu The observations are summarised in Table3.

MNRAS 000,1–30 (2019) 6 C. Fontanive et al.

Table 3. Summary of our new observations.

Target Observation Date Telescope / Instrument Filter Field of View Pixel Scale Previous Observations

WASP-18 September 4, 2017 VLT / NACO L’ 2800 × 2800 000. 027 Ngo et al.(2015) HD 162020 September 6, 2017 VLT / NACO L’ 2800 × 2800 000. 027 Eggenberger et al.(2007) BD+24 4697 September 6, 2017 Gemini North / NIRI Ks 2200 × 2200 000. 022 − HD 77065 December 12, 2017 Gemini North / NIRI Ks 2200 × 2200 000. 022 − HD 134113 June 22, 2018 WIYN / NESSI 562nm, 832nm 4.600 × 4.600 000. 040 − HD 160508 June 24, 2018 WIYN / NESSI 562nm, 832nm 4.600 × 4.600 000. 040 −

3.1.1 VLT / NACO observations We followed standard procedures for near-infrared data reduction, using the Gemini NIRI IRAF package and our dedicated IDL We obtained images in the L0 filter (3.8 µm) using the AO-assisted routines. A sky frame was constructed by taking the median of the imager NACO at VLT (Lenzen et al. 2003; Rousset et al. 2003) for dithered images, masking the regions dominated by the target’s sig- HD 162020 and WASP-18 (programme 099.C-0728, PI Fontanive). nal. The individual images were then sky-subtracted and divided by These new data were acquired with the aim to confirm or refute a normalised flat field, and bad pixels were replaced by a median a candidate reported in Eggenberger et al.(2007) around the for- calculated over their good neighbours. For all images, field distor- mer target, and to achieve deeper detection limits than in currently tion was corrected as described in Lafrenière et al.(2014). No can- available imaging data of the latter object (Ngo et al. 2015; see Ap- didate companion was identified around either target. pendixA). The observing setup included the L27 camera, and the data were taken in the pupil tracking mode, where the telescope pupil is held fixed, and the field rotates. Each target was observed 3.1.3 WIYN / NESSI observations using a three-point dither pattern, designed to avoid a bad quad- rant of the NACO detector. We used short integration time (0.2 s) We acquired observations of HD 160508 and HD 134113 with in order not to saturate the primaries, allowing photometric and as- the WIYN 3.5-m telescope at Kitt Peak National Observa- trometric calibrations. tory (KPNO). We used the NASA-NSF Exoplanet Observational Standard near-infrared data reduction techniques were applied Research (NN-Explore) Exoplanet and Stellar Speckle Imager using our custom IDL routines, including sky subtraction, flat- (NESSI) in diffraction-limited speckle imaging mode. NESSI is fielding and bad-pixel correction. Some of the frames were affected based on an upgraded design of the Differential Speckle Survey In- by the horizontal additive noise pattern, that sporadically appeared strument (DSSI; Horch et al. 2009, 2012). Each target was observed in the NACO data, and was variable in intensity and time. The simultaneously in two cameras, with a filter centered on 562 nm (r- pattern was removed following the procedure described in Huß- narrow) on the blue channel and a bandpass at 832 nm (z-narrow) mann et al.(2012). Individual frames were de-rotated according to on the red channel. The standard NESSI observing strategy was the parallactic angle, and finally stacked together. We retrieved in followed, with typical integration times of 40 ms (see Scott et al. our final images the unconfirmed candidate companion around HD 2018). Data were reduced by the KPNO speckle reduction pipeline 162020 reported by Eggenberger et al.(2007) and were able to re- that generates reconstructed images and contrast limit curves for fute the bound nature of this source based on our new data. The each observation (Scott et al. 2018). We did not identify any candi- detailed analysis of the rejected candidate is presented in Section date companion in the obtained data around these two targets. 3.3. No companion was detected around WASP-18 within the field of view of our images. 3.2 Achieved sensitivities We estimated the limiting sensitivities reached around our observed 3.1.2 Gemini North / NIRI observations targets in order to establish the full range of detectable companions We acquired images in Ks band (1.95−2.30 µm) using the Gem- covered by the obtained data. For the VLT / NACO and Gemini ini Near-Infrared Imager (NIRI; Hodapp et al. 2003) instrument at North / NIRI data, detection limits were determined from the final the Gemini North telescope for BD+24 4697 and HD 77065 (pro- images described above. The 5-σ noise curves were calculated as a gramme GN-2017B-Q-40, PI Fontanive). Targets were observed in function of radius by computing the standard deviation in circular the standard imaging mode, using the Gemini North adaptive optics annuli with 1-pixel widths, centred on the primary targets. Noise (AO) system ALTAIR (Herriot et al. 2000) to obtain diffraction- levels were then converted into magnitude contrasts by dividing limited images with the f/32 camera. Both our target were bright by the peak pixel values of the targets (which were not saturated), enough to be used as natural guide stars. The observing strategy and converting the obtained flux ratios into magnitude differences adopted was similar to the one described in Lafrenière et al.(2008) in the considered filters. The contrast curves generated from the and Daemgen et al.(2015). Each target was observed at five dither custom KPNO pipeline in the 832 nm filter were considered for positions to allow for sky subtraction and bad pixel correction. At the WIYN data, as the redder filter is better suited for the detec- each dither position we acquired one non-saturated short exposure tion of warm, low-mass companions. The achieved magnitude con- (divided into many coadds) in high read noise mode, followed by a trasts are presented in Figures1,2 and3 for our NACO, NIRI and longer exposure in low read noise mode. This prevents our obser- NESSI observations, respectively. The hydrogen-burning limits are vations from being limited by the high read out noise, resulting in shown for the first two data sets, showing that we are sensitive to a high observing efficiency and a large dynamic range, providing substellar companions around these stars. We did not reach the stel- sensitivity at both small and large separations. Our targets were not lar/substellar boundary in the WIYN observations and are only able saturated, even in the deeper exposures. to detect low-mass stellar companions in these data.

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 7

Figure 4. Common proper motion analysis of the faint companion around HD 162020, originally identified by Eggenberger et al.(2007). The black Figure 1. Achieved sensitivities showing the 5-σ magnitude differences in solid line represents the motion of a background object relative to the pri- L0 as a function of angular separation for our two targets observed with mary, computed using the proper motion and parallax measurements of HD VLT/NACO. The dashed lines indicate the magnitude differences corre- 162020 from Gaia DR2. The blue and red circles mark the relative posi- sponding to the hydrogen-burning limit for each target. tions of the components in our new NACO observations and in the data from Eggenberger et al.(2007), respectively. The red crosses indicate the expected positions of a background source at the dates of the observations used by Eggenberger et al.(2007). The relative motion of the candidate over the available epochs is not consistent with a comoving pair.

3.3 Refuted candidate around HD 162020 HD 162020 had previously been observed with NACO as part of the survey conducted in Eggenberger et al.(2007). Eggenberger et al. (2007) reported two point sources within 500from the star, one of which was found by the authors to clearly be a background source. They found that the second candidate, at 400. 98 ± 000. 03, was more likely unbound than bound, although the low significance level of this result led them to report the companionship of this candidate as inconclusive based on their data alone. Both sources are retrieved in our new NACO images (Section3). The positions of the detected sources were extracted using the StarFinder PSF-fitting algorithm (Diolaiti et al. 2000), employing an empirical PSF extracted from Figure 2. 5-σ magnitude differences achieved in Ks for our two targets observed with Gemini North/NIRI, showing the corresponding hydrogen- the primary. burning limits (dashed lines). To calibrate the pixel scale and the True North (TN) of the detector, we used the astrometric calibrator system θ1 Orionis C, observed on October 6 2017. Using the same procedure as described in Chauvin et al.(2012), we obtain the pixel scale of 27.10±0.05 mas, and the TN position of −0.45±0.10 deg. However, as previously pointed out by Eggenberger et al.(2007) and Chauvin et al.(2012), additional systematic errors might be present in the determination of the TN of the NACO detector in a case where different sets of calibrator stars were used between the epochs. Since we do not know which calibrators were used to derive astrometry in the previous epochs by Eggenberger et al.(2007), we add 0 .5 deg to the TN error budget (G. Chauvin, priv. comm.). With a baseline larger than a decade between the observations used in Eggenberger et al.(2007) and ours, and given the proper motion of the primary from Gaia DR2, we were able to refute the bound nature of this companion. Figure4 shows the relative positions of the primary and candidate at the various epochs available, together with the expected motion of a background object. The plot clearly demonstrates that the candidate does not share com- Figure 3. 5-σ magnitude differences achieved in the 832 nm filter for our mon proper motion with the primary. The fact that the relative po- two targets observed with WIYN/NESSI. These observations do not reach sitions at the old epochs of observation are not consistent with the the hydrogen-burning limit. primary, nor with the expected background positions, suggests that the source has a non-negligible proper motion of its own.

MNRAS 000,1–30 (2019) 8 C. Fontanive et al.

Figure 5. Orbital properties of the inner companions in our sample showing the systems that are known to be binaries or higher-order multiples (stars), Figure 6. Architecture of all binary or higher-order multiple systems found those with a candidate companion (triangles) and stars that are apparently in our sample, following the order of the targets shown in Tables4 and5. single (circles). The blue circles represent the inner brown dwarfs and planets, with sym- bol sizes proportional to their masses. Red circles show the positions of all known confirmed (filled symbols) and candidate (open symbols) wide 4 SEARCH FOR WIDE COMPANIONS companions, with radii proportional to the mass ratios of these outer companions to the planet hosts. Separations of inner companions correspond to We searched for wide companions to the 38 systems from our core semi-major axes, while observed projected separations are displayed for the sample using our new data and published direct imaging observa- wide binary companions. tions, as well as the Gaia DR2 catalogue. A total of 16 objects were found to have at least one wide stellar or substellar companion confirmed to be comoving, listed in Table4. Another 7 candidate (see Figure A1). We list all confirmed multiples in Table4. Three of companions are reported in the literature around 4 of our targets these systems, HD 87646, HD 41004 and HD 178911, were identi- and are presented in Table5. One of the targets with a reported fied as binaries in the Tycho-Hipparcos catalogues and for the latter candidate is already a confirmed wide binary (HD 89744). Figure two systems, our planet-host stars correspond to the fainter compo- 5 displays the properties of the inner planets and brown dwarfs, nent of the binary system. showing the positions in the planet period-mass space of confirmed The remaining 6 targets are mentioned to have unconfirmed binaries (star symbols), targets with a candidate companion (trian- candidate companions and are discussed in Appendices A2 and gles) and apparently single objects (circles). In Figure6 we present A3. We discarded the 3 point sources reported by Moutou et al. the architecture of each identified hierarchical system, plotting the (2017) around HD 168443, which are highly likely to be back- semi-major axes of the inner companions in blue and the projected ground contaminants given the crowded galactic latitude of the separations of detected wide binary components in red, with sym- target (Moutou et al. 2017; see discussion of HD 168443 in Ap- bol sizes proportional to the planetary masses and binary mass ra- pendix A3). In Section 3.3, we showed that the faint candidate re- tios, respectively. ported around HD 162020 by Eggenberger et al.(2007) does not share common proper motion with the primary and thus rejected this candidate. We were also able to refute the candidate compan- 4.1 Literature search and imaging surveys ion reported around XO-3 by Bergfors et al.(2013) and Wöllert We conducted an extensive literature search to compile available & Brandner(2015) based on the inconsistent parallax and proper observations of all objects in our sample and gather existing knowl- motion of this source and XO-3 in Gaia DR2. This leaves 3 tar- edge about the multiplicity of our targets. We present our findings gets with unconfirmed candidates, namely, 70 Vir (two candidates), for each individual target in AppendixA, providing detailed infor- EPIC 219388192 (three candidates) and KELT-1 (one candidate). mation about every companion, candidate or confirmed, reported A candidate companion is also reported by Roberts et al.(2011) around our targets in imaging surveys or catalogues, as well as null- around HD 89744, already known to be a confirmed wide binary detections. A total of 30 targets are mentioned in the literature in (Mugrauer et al. 2004). These final 7 candidate companions re- the context of a search for wide companions (with or without de- tained for this study are presented in Table5. tections), to which we add 4 of our 6 observed targets that had no We are not able to make any clear statement on the physi- previously reported observations (see Section3). cal association of these candidates based on the available data. We Of these 34 objects, we found 21 targets with reported detec- can however make a statistical argument on the chance of finding tions in the literature. Among those, 15 are confirmed bound sys- an unrelated background source at close angular separation from tems (11 Com AB, 30 Ari ABC, τ Gem AB, υ And AB, AS 205 the primaries. For each source, we used the Trilegal galaxy models ABC, HAT-P-20 AB, HD 41004 AB, HD 87646 AB, HD 89744 (Vanhollebeke et al. 2009) to calculate a probability of the observed AB, HD 114762 AB, HD 156846 AB, HD 178911 ABC, Kepler- candidates being true companions. This was done by estimating 13 AB, NLTT 41135 AB and WASP-14 AB), which we detail in the surface density ρ of background sources expected to be found Appendix A1. 14 of these binaries or higher-order multiple were within 300from the primary targets, given the galactic latitude and demonstrated to form physical pairs in the literature and we con- longitude of the objects and the depth and wavelength of the ob- firm the true companionship of the τ Gem AB system in this work tained observations. From Brandner et al.(2000), the probability

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 9 this work Mugrauer et al. ( 2004 ); GDR2 Ma et al. ( 2016 ); Hipparcos-Tycho, WDS; CCDM Patience et al. ( 2002 ) Tamuz et al. ( 2008 ); WDS; GDR2 Zucker et al. ( 2003 ); WDS; CCDM; Hipparcos-Tycho WDS; Bakos et al. ( 2011 ) Lowrance et al. ( 2002 ); GDR2 CCDM; WDS; CCDM; GDR2 B-C pair: Riddle et al. ( 2015 ); Roberts et al. ( 2015 ) References Irwin et al. ( 2010 ); GDR2 54 14 13 07 07 58 18 24 43 14 A-BC pair: Guenther et al. ( 2009 ); GDR2 09 . . 35 . . 24 B-C pair: Santerne et. al. ( 2012 ); Johnson et al. ( 2014 ) 26 A-BC pair: Szabó et al. ( 2011 ); Shporer et al. ( 2014 ); GDR2 . 43 BC pair: Eisner et06 al. ( 2005 ) 14 A-BC pair: Ghez et al. ( 1993 ); Prato et al. ( 2003 ); GDR2 . . ) 05 AC-B pair: Tokovinin et al. ( 2000 ); Hipparcos-Tycho; GDR2 66 A-C pair: McAlister et al. ( 1987 ) ...... 28 05 56 . . 1 0 1 0 0 0 0 0 0 . . . 24 0 0 0 0 0 0 1 0 0 . 0 0 − 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± δ ± ± ± ± µ 20 84 90 90 14 08 72 98 28 19 68 . . 78 . 08 . . 05 . 17 23 89 . . 90 13 ...... 17 64 65 . . . . 26 . . 23 26 48 15 10 15 27 15 96 138 140 282 143 132 281 382 383 − − − − − − − − − − − − − − − − − 71 2 10 70 52 13 32 88 08 89 11 98 49 ...... 48 59 55 . . ) (mas yr 14 08 18 24 07 47 20 . 19 . . 18 04 195 . . . . 69 207 1 0 0 0 0 0 0 0 . . . . . 0 0 . . 0 0 1 0 0 0 0 − 0 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± α ± ± ± ± ± ± ± 93 29 10 62 25 08 24 09 57 . 49 ...... 52 88 . 86 41 51 67 . 99 48 45 . . 40 95 18 62 ...... 4 4 3 9 7 31 41 82 109 108 119 120 586 137 129 172 172 − − − − − − − − − − − − − − − − − 95 27 05 42 03 57 26 38 76 04 09 35 07 136 22 05 141 07 12 162 16 153 . 32 . . 05 ...... 78 11 . . 08 19 10 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± π µ ± ± ± ± ± ± ± 96 85 86 92 43 38 04 23 05 21 57 13 71 36 27 11 . . 88 . . 56 ...... 40 91 . . . 09 38 82 ...... G 12.2 9 12.80 ...... ) (mag) (mas) (mas yr

0.7 0.57 6.9 1.16 6.96 22 = 8.3 1.03 7.87 24 8.88.3 0.7 8.38 1.12 24 7.93 7 6.8 1.9 6.57 20 i 11 0.8 9.42 ...... 11.0 0.6 ...... 12.5 0.4 ...... 4.5 2.3 3.95 8 6.5 1.32 6.38 22 4.8 2.02 4.37 10 4.9 1.5586.2 5.59 25 0.83 7.14 24 5.59.4 1.38 0.59 6.36 12.19 20 21 9.3 0.756 10.99 14 9.2 0.22 13.40 6 8.1 1.086 12.37 7 9.4 0.2 12.51 74 3.2 1.31 3.90 74 12.9 = = = = 14.9 0.08 19.66 25 13.8 0.09 ...... 10.2 13.1 0.164 14.94 29 12.4 0.21 13.97 29 11.2 0.5 ...... 10.48 1.68 10.37 1 10.35 1.72 10.55 2 ======V J J J J J J J J J 7.1, V V V i z z J J J V Hp Hp Hp Hp V V = Hp Hp V K1 V GV M0 V + + + F8 V F8 V F7 V F9 V G1 V G5 V K3 V A5 V K5 V M2 V K2 III G8 III G1 IV M5.1 V B M4.5 V B K0 V BC ... M1 V B B L0 V BKV B M9 V B M4 V B B B BMV A A A A A A AA K1 V A M4.2 V A A A BC A BC K7 AC G1 ) (AU) (mag) (M 00 ( 38.2 1520 A F5 V 0.536 22 Conﬁrmed common proper motion systems. And 55 750 Gem 1.9 187 υ τ System Separation11 Com Comp. 9.1 SpT30 Ari 850 Photometry Mass HD 89744 63HD 114762 2460 3.2 140 HD 156846 5.1 250 NLTT 41135 2.4 55 HD 178911 16.1 790 HD 87646 0.26 20 HD 41004 0.54 23 HAT-P-20 6.86 500 Kepler-13 1.15 610 AS 205 1.3 166 Table 4.

MNRAS 000,1–30 (2019) 10 C. Fontanive et al. -magnitudes, parallaxes G ; GDR2. this work This work; GDR2 References A-B pair: Ngo et al. ( 2015 ) ) 06 18 AB-C pair: 03 40 1 . . . . − 0 0 2 0 δ ± ± ± ± µ 95 15 38 60 . . . . 6 6 − − ) (mas yr 98 18 03 20 06 20 1 . . . . − 0 1 0 0 α ± ± ± ± 65 24 24 97 . . . . 03 29 10 27 52 23 02 25 . . . . 1 0 0 0 ) (AU) (%) π µ ± ± ± ± 00 5.62 219 99.17 likely bound Roberts et al. ( 2011 ) 2.86 52 99.67 likely bound Roberts et al. ( 2011 ) 42.7 848 76.91 inconclusive Pinﬁeld et al. ( 2006 ) 07 14 08 43 0.588 154 99.95 likely bound Siverd et( 2012 ) al. . . . . )(

0.1 7.538 2224 4.65 likely background Nowak et al. ( 2017 ) 0.1 5.988 1769 14.33 likely background Nowak et al. ( 2017 ) 0.2 0.08 0.08 0.07 < < G 17.32 6 20.92 9 12 019 1.335 – – – – ) (mag) (mas) (mas yr . .

0 0 ± ± 0.092 0.280 59 666 1.01 – – – – . . 437 5 . 2 80 1.07 – – – – 057 032 . 16 . . . 9 10 . 2 1 = 3 0 0 0 = = 24 0.52 0.082 24 99.99 likely bound Curtis et al. (private communication) ± ± ± ± = 2 1.558 – – – – . K ± . 2 9.3 1.20 9.17 8 8.9 1.35 9.65 6 4 K K ∆ J 5 14.1 0.33 ...... 13 , , , = , = 84 = = . 663 087 = J V 11 . . J = 10 (mag) (M 98 I . 7 7 K . 15 030 734 I = 0 . . 3 ∆ = = ∆ 0 I = ± 10 = ∆ ± J H H I = 90 ∆ ∆ . 5 V H 534 / . = 9 = H F5 V K5 V M7.5 V ∆ F6 IV H B B ... A A M8 V M8 V M5 V GV F7 V F5 V G5 V > > > C ... B M4-5 V C BMV CLV B A A D A A ) (AU) (mag) (M 00 ( 11.5 1900 C 1.45 300 (Continued.) Candidate binary companions. The top component of each system (marked in bold) is the planet host considered in the sample studied here. Spectral types and masses in italic were derived in this work. The top component of each system (marked in bold) is the planet host considered here. HD 89744 KELT-1 EPIC 219388192 System SeparationWASP-14 Comp. SpT PhotometryNotes. Mass System70 Vir Comp. SpT Photometry MassNotes. Separation Prob. Companionship References and proper motions come from theCCDM: Gaia Catalog DR2 of catalogue. Components ofGDR2: Double Gaia and Data Multiple Release stars, 2, DommangetGaiaWDS: & Collaboration the Nys ( 2000 ). et Washington Double al. ( 2018 ). Star Catalog, Mason et al. ( 2001 ). WASP-18 26.7 3300 Table 4. Table 5.

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 11

P(Θ, m) of detecting one or more background stars within an angu- time spans are required to ensure that the observed proper motions lar separation Θ (in arcsec) and down to a limiting magnitude m is correspond to the true barycentric motion (e.g. the Tycho-2 cata- then given by: logue which uses data from over a century timescale). The apparent 2 changes in proper motions between short and long-term measure- P(Θ, m) = 1 − e−πΘ ρ(m). (1) ments can be as large as several tens of mas yr−1, based on the com- The probability of an observed candidate being physically associ- ponents masses, binary separation, orbital phase and parallax of the ated to the primary is then given by the complement of the chance system (Shaya & Olling 2011). These changes in proper motion of alignment, that is, 1 − P(Θ, m). The resulting probabilities are may even be exploited as a way to search for hidden companions, listed in Table5 for each candidate. The two faint candidates identi- as was done by Makarov & Kaplan(2005) with the Hipparcos and fied beyond 700from EPIC 219388192 by Nowak et al.(2017) were Tycho-2 catalogues. Despite the excellent precision of Gaia DR2, found to likely be background sources, with probabilities <15% of the catalogue is based entirely on data collected between July 2014 being physically associated. With the exception of the wider candi- and May 2016, spanning a period of only 22 months, and the same date around 70 Vir, most other candidates were found to have very problem as for Hipparcos is encountered. high probabilities of being bonafide companions: the close can- While these effects are reduced at very wide separations, fur- didates around 70 Vir, EPIC 219388192, HD 89744 and KELT-1 ther complications can also arise from the presence of an unre- have >99% probabilities of being bound. While additional observa- solved binary. Shaya & Olling(2011) estimated that a tight system tions will be required to confirm their true companionship through separated by a few AU could induce proper motion fluctuations of common proper motion analyses, these objects are therefore highly several mas yr−1 on the primary, orders of magnitude larger than the likely to be true companions. errors on Gaia DR2 measurements. Close binaries not resolved in Finally, a total of 9 targets from our core sample are men- Gaia are treated as single objects in the second Data Release, which tioned as single objects in the literature and have reported null de- can lead to specious astrometric solutions (Arenou et al. 2018). A tections from direct imaging surveys (CI Tau, HAT-P-2, HD 5891, third component at a wide separation around an unresolved binary HD 33564, HD 104985, HD 156279, HD 180314, HD 203949 and is therefore likely to show somewhat different astrometric parame- WASP-18; see Appendix A4). We add to these objects 4 of our ob- ters (proper motion and parallax) compared to its comoving, unre- served targets that had no previous observations (BD+24 4697, HD solved primary. 77065, HD 134113 and HD 160508) and around which we did not As a result, we adopted rather loose selection criteria to search find any companion. No published data were found in the literature for comoving companions to the objects in our sample using the for the targets 4 UMa, 59 Dra, HD 39392 and HD 112410, which Gaia DR2 catalogue. We considered the relative differences in par- we were not able to observe either. allax π, i.e. ∆π/π0 ≡ |(π0 − πi)/π0|, where the subscript 0 corresponds to our science target and i to other Gaia sources. We then defined similar relative differences for the proper motion compo- 4.2 Companions in Gaia DR2 nents, µα∗ and µδ. To account for the uncertainties in the Gaia mea- Raghavan et al.(2010) found that the period distribution of binary surements, we generated, for each pair of objects, 105 parallaxes companions to nearby FGK stars is approximately a Gaussian in and proper motions drawn from Gaussian distributions centred on the logarithm of the period, with a broad peak around 300 yrs (∼50 the measured values, with a standard deviation set to the Gaia un- AU), and a 1-σ Gaussian interval spanning from 2 to 1500 AU, certainties. We then calculated 105 corresponding fractional differ- in reasonable agreement with previous studies by Duquennoy & ences in π, µα∗ and µδ and set the final relative differences and as- Mayor(1991). Most of the imaging data considered here only allow sociated uncertainties to the mean and standard deviation of the for the detection of companions out to several hundred AU. Hence output distributions. a significant number of wider companions could remain outside the We selected sources that were consistent with relative differ- field of view of these observations and be missed by direct imaging ences of less than 20% in parallax and in at least one of the two surveys. At these wide separations, outer companions are expected proper motion components (including the correct direction), with a to be massive (i.e. a stellar binary) in order to be able to affect the maximum relative discrepancy of 50% in the other proper motion formation or evolution of close-in planets or brown dwarfs. Such component. We searched for such companions in the Gaia DR2 wide stellar companions are expected to be found in the Gaia Data catalogue for all targets in our sample, out to angular separations Release 2 (DR2; Gaia Collaboration et al. 2016, 2018), which may corresponding to projected separations of 104 AU. We found a to- thus be used to search for widely-separated comoving components tal of 11 systems fulfilling the above selection criteria, 9 of which to the objects in our sample. We therefore searched for Gaia DR2 were previously known systems. These systems are listed in Table sources with parallaxes and proper motions consistent with those 4, in which we give the Gaia DR2 parallaxes and proper motions of our targets, to complement our direct imaging search for wide for each binary component. The characterisation of the two newly- companions in Section 4.1. identified Gaia systems, WASP-14 AB-C and WASP-18 A-B, is The recent release of the Gaia DR2 catalogue provides detailed in Sections 4.2.1 and 4.2.2 below. unprecedentedly-precise astrometric measurements on the paral- In Figure7 we plot the relative di fferences in parallax and laxes and proper motions of stars. However, these highly-precise proper motion (RA and Dec coordinates) between the components measurements must be considered and handled with caution in the of all identified Gaia binaries. The shaded area represents our ar- context of a search for common proper motion systems. Shaya bitrary cut at 20% in the relative differences in parallax and proper & Olling(2011) accurately pointed out that astrometric missions motion. The 9 previously known systems are marked in blue and spanning a few years only (e.g. Hipparcos, 3.5 year baseline) cap- our the two new systems discovered here are shown in red. The ob- ture the reflex motions of multiple systems in their kinematics tained values and their associated uncertainties are given in Table6 measurements. Indeed, the components of a binary wobble around and are all consistent with our chosen constraints at the 1-σ level. the centre of mass of the system and a short-term proper motion In AppendixB we examined those systems more carefully, measurement is highly likely to reflect this orbital motion. Longer as well as other known binaries in our sample, in order to assess

MNRAS 000,1–30 (2019) 12 C. Fontanive et al.

Table 6. Relative differences in parallax and proper motion, with their as- 4.2.1 WASP-14 sociated errors, between the components of all Gaia binaries. Fractional WASP-14 A is already known to have a 0.33 M bound companion differences are calculated relative to the first component listed for each sys- at 300 AU (Wöllert & Brandner 2015; Ngo et al. 2015) as discussed tem (single or binary), and our science targets always correspond to the first component given. The two new binaries identified in this work are marked in Appendix A1 (see Table4). This companion is not detected in 00 in bold. Gaia due to the small angular separation (1. 45) and large magnitude difference (∆J=5.2 mag) of the WASP-14 A-B system (see discussion in AppendixB). We report a new companion to this sys- System ∆π/π0 ∆µα∗/µα∗,0 ∆µδ/µδ,0 (%) (%) (%) tem, WASP-14 C (Gaia DR2 1242084166679297920), at a separation of 1100. 5397±000. 0001 and a position angle of 4.5827±0.0003 11 Com ...... A-B 10.74 ± 1.92 1.05 ± 0.30 1.67 ± 0.33 deg. The measured angular separation corresponds to a wide pro- ± ± ± 30 Ari ...... BC-A 1.03 0.39 3.21 0.11 42.23 1.81 jected separation of 1900 AU at the distance of WASP-14 (see Ta- υ And...... A-B 0.48 ± 0.38 0.10 ± 0.19 0.26 ± 0.14 ble4). WASP-14 AB and C have measured Gaia DR2 parallaxes in AS 205 ...... A-BC 18.41 ± 1.55 27.25 ± 7.14 13.83 ± 1.66 HD 89744. . . . .A-B 0.43 ± 2.24 0.53 ± 0.58 1.40 ± 0.75 excellent agreement, with a relative difference <1%. The relative HD 156846. . . .A-B 2.44 ± 1.70 5.46 ± 0.51 7.23 ± 0.41 discrepancies in proper motion are slightly larger but still in very HD 178911 . . B-AC 17.02 ± 1.58 33.40 ± 1.20 5.73 ± 0.34 good agreement: 4.34% in µα∗ and 11.51% in µδ. Given the consis- Kepler-13. . . .A-BC 8.61 ± 5.46 11.39 ± 6.55 4.85 ± 2.38 tent parallax and small offsets in proper motion, we conclude that NLTT 41135 . . B-A 0.55 ± 0.51 5.44 ± 0.18 0.26 ± 0.11 the two objects are comoving and form a physical pair. Comparing WASP-14 . . . AB-C 0.98 ± 1.17 4.34 ± 0.71 11.51 ± 2.69 the placement of WASP-14 in Figure7 to the other Gaia binaries in WASP-18. . . . . A-B 16.85 ± 14.60 6.30 ± 5.94 10.78 ± 9.03 our sample also reinforces the idea that WASP-14 is a true binary and confirms our intuition that systems with an unresolved component tend to show larger disparities in their observed short-term proper motions. WASP-14 C has a Gaia G-band magnitude of 17.32 mag, for a magnitude difference of ∆G=7.67 mag with the unresolved WASP- 14 AB primary. Photometry in the blue (GBP) and red (GRP) filters of Gaia indicate fairly red colours for this object, with GBP −GRP = 2.67 mag. This suggests a mid-K to early-M main sequence star according to the Gaia DR2 HR diagram analysis in Gaia Collabo- ration et al.(2018). The new companion to WASP-14 is also found in the 2MASS catalogue, with magnitudes of J = 14.297 ± 0.054, H = 13.801 ± 0.049 and Ks = 13.592 ± 0.058. According to Schmidt-Kaler(1982), the 2MASS colours correspond to a K5 V spectral type. This implies a bolometric correction of BCK of ∼2.3±0.1 mag (Masana et al. 2006). From these values, we calculated a bolometric luminosity and used the BT-Settl models (Allard et al. 2012) to infer a mass for WASP-14 C. Adopting a distance based on the Gaia DR2 parallax of the target and the age of the system given in Table2, we derived a mass 0 .280 ± 0.016 M for the newly discovered stellar component of the triple system WASP-14, making this companion the lowest-mass component of the system. With its low mass and extremely wide separation, WASP-14 Figure 7. Binaries identified in Gaia DR2, showing the relative difference C is unlikely to have played a role in the formation or evolution of in parallax (x-axis) against the relative differences in proper motion (y-axis, both RA and Dec components) between the science target and the selected the 7.8 MJup planet on a 2.2 day orbit around WASP-14 A, as the companion. Our selection criteria correspond to the shaded area (see text). closer and more massive WASP-14 B component would have had a Systems marked with stars rather than circles indicate binaries that have a much stronger influence (if any) on the inner substellar companion. component known to be unresolved in Gaia. The two new binaries identified In Figure8 we show the relative positions of WASP-14 AB and C in this work are marked in red. We do not plot error-bars for clarity of the from the 2MASS and Gaia DR2 catalogues, confirming over a ∼20- figure but show them in Table6 instead. year baseline that the companion is indeed comoving. The WASP- 14 A-B pair is unresolved in both 2MASS and Gaia. our selection criteria. We found that our selection method success- 4.2.2 WASP-18 fully identified all known binaries that were recoverable given the sensitivity and completeness of Gaia DR2, and consider that we We report here the detection of a faint comoving companion at unlikely missed additional binaries present in the Gaia DR2 cata- 2600. 728±000. 001 (∼3300 AU projected separation) and a position logue. Based on the location of systems with a known unresolved angle of 200.520±0.001 deg, found in the Gaia DR2 catalogue component in Figure7 (marked with stars), we conclude that most (Gaia DR2 4931352153572401152), outside the field of view of binaries should have relative discrepancies of <10% in all astro- our VLT/NACO data (see Section3). The same companion was in- metric parameters (π, µα∗ and µδ), while systems agreeing to within dependently identified by Csizmadia et al.(2019) at the same time. 20% in parallax and in one of the proper motion coordinates are The source has a Gaia magnitude G = 20.92 mag, and a G − GRP likely to be hierarchical systems with an unresolved component colour of 1.79 mag. We do not consider the GBP photometry of this (see AppendixB for details). source because of the known poor quality of fluxes in this bandpass

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 13

2MASS 200.5 2MASS 7 Gaia DR2 Gaia DR2 200.4

6 200.3

200.2

5 200.1 Position Angle (deg) Position Angle (deg)

200.0

4 199.9 11.40 11.45 11.50 11.55 11.60 26.3 26.4 26.5 26.6 26.7 26.8 Separation (arcsec) Separation (arcsec)

Figure 8. Common proper motion analysis of the WASP-14 AB-C system. Figure 9. Same as Figure8 for the WASP-18 A-B system. The Gaia DR2 The black solid line represents the motion of a background object relative to epoch is again 2015.5 and the 2MASS observations date from August 1999, the primary, WASP-14 AB, computed using the proper motion and parallax providing a 16-year time span, clearly demonstrating that the two objects measurements of our science target the Gaia DR2 catalogue. The blue cross share common proper motion. and red square mark the measured position of the components in Gaia DR2 and in 2MASS, respectively. The red circle indicates the expected position of a background source at the date of the 2MASS observations (June 1997). measurement errors we found fractional differences in parallax and The Gaia DR2 epoch is 2015.5, providing an 18-year baseline between the proper motion between WASP-18 A and B of 16.85 ± 14.60 %, two epochs available. As expected from the Gaia astrometry of the system, 6.30 ± 5.94 % and 10.78 ± 9.03 %, respectively. While these er- the relative motion of the companion between the two epochs is consistent rors are all very large and comparable to the obtained value for ∆π, with a comoving pair. ∆µα∗ and ∆µδ, they are still consistent with our selection criteria at the 1-σ level. In particular, the 1-σ intervals in the fractional difference of both proper motion components remain within 20%, which for red sources with G ∼ 20−21 mag (Gaia Collaboration et al. strongly indicates that both objects are travelling in the same di- 2018). At the distance of WASP-18, this places the companion on rection. With the close angular proximity of the two sources (2600) the M/L-dwarf sequence in the HR diagram for Gaia DR2 sources and parallax measurements consistent with each other, we conclude presented in Gaia Collaboration et al.(2018). that WASP-18 A and B form a physically associated pair. We searched for the same object in the 2MASS catalogue This is further supported by the red colours of the companion, and retrieved a source at the same relative position (2600. 71±000. 15 its placement on the Gaia HR diagram, and the consistent relative and 200.36±0.08 deg), in excellent agreement with the consistent positions of the two components over 16 years between 2MASS proper motions of the two sources in Gaia DR2. Figure9 shows the and Gaia DR2 (Figure9). The possible o ffset in the Gaia DR2 relative positions of WASP-18 A and B at the time of the 2MASS parallaxes could indicate that one of the components is an unre- and Gaia DR2 observations, together with the expected motion of solved binary, as we found in our analysis that systems with an a background source between the two epochs. The two objects are unresolved component tend to show higher inconsistencies in the clearly found to be comoving based on this analysis, confirming short-term astrometric measurements available in Gaia DR2 (e.g. our findings from the Gaia DR2 catalogue. The 2MASS photome- AS 205 A-BC and HD 178911 AC-B have relative differences in try of WASP-18 B is J = 16.289 ± 0.096, H = 15.513 ± 0.083 and parallax close to 20%). Alternatively, this discrepancy could also Ks = 15.146 ± 0.121. Using the relation between spectral type and reflect a much wider binary separation, along the line of sight, with MJ from Filippazzo et al.(2015), we infer a spectral type of M7.5 one component located at a slightly larger distance. Further obser- for the companion, in agreement with our rough estimate from the vations will be required to reduce the uncertainties in the parallax Gaia HR diagram. We used this spectral type to estimate a bolomet- of WASP-18 B, which will be available in future data releases of ric correction BCJ from the relations in Filippazzo et al.(2015) for the Gaia mission. Despite these large uncertainties, we are never- field objects, and derived a corresponding bolometric luminosity. theless able to confirm that WASP-18 A and B form of a common We finally interpolated the obtained luminosity into the BT-Settl proper motion pair, in which the fainter component is a low-mass evolutionary models (Allard et al. 2012) at an age of 0.90 ± 0.20 M dwarf close to the stellar/substellar boundary. Gyr (Bonfanti et al. 2016), to obtain a mass of 0.092 ± 0.003 M for WASP-18 B. WASP-18 B is our Gaia source with the largest uncertainties 5 DETECTION LIMITS in its parallax and proper motion measurements, with significant errors of 1.52 mas in π, 1.98 mas in µα∗ and 2.40 mas µδ. This is In order to take into account observational biases and survey sensi- due to the fact that this source only has 169 observations in the tivities in our analysis, we gathered and generated detection limits Along-Scan direction and none in the Across-Scan direction. In for each target in our sample. We searched for existing contrast comparison, WASP-14 A has 412 observations in both the Along- curves from the literature for the targets with previous observations scan and Across-scan directions, allowing a much higher precision (Section 5.1) and derived Gaia detection limits for all targets in our on its astrometric measurements (0.02−0.03 mas). Propagating the sample (Section 5.2). Combining those with our own sensitivity

MNRAS 000,1–30 (2019) 14 C. Fontanive et al.

Table 7. Detection limits used.

Object ID Facilities / Instrument Filter Phot. (mag) Limits units Curve ﬂag Data Set

30 Ari B Keck / NIRC2 J 6.080 ± 0.02 ∆mag 0 Kane et al.(2015) 70 Vir AEOS I 3.98 mag 2 Roberts et al.(2011) a 2MASS J 3.80 mag 2 Pinfield et al.(2006) υ And AEOS I 3.35 mag 2 Roberts et al.(2011) a Lick / LIRC II K0 2.86 ∆mag 0 Patience et al.(2002) BD+24 4697 Gemini North / NIRI Ks 7.474 ± 0.023 ∆mag 0 This paper CI Tau Subaru / HiCIAO H 8.43 ± 0.04 ∆mag 0 Uyama et al.(2017) EPIC 219388192 Keck / NIRC2 K 10.666 ± 0.0216 ∆mag 0 Curtis et al. (private communication) Subaru / IRCS H 10.734 ± 0.021 flux ratio 0 Nowak et al.(2017) HAT-P-2 Calar Alto / AstraLux z0 9.506 ± 0.001 ∆mag 1 Bergfors et al.(2013) HAT-P-20 Keck / NIRC2 K 8.601 ± 0.019 ∆mag 1 Ngo et al.(2015) HD 5891 Calar Alto / AstraLux i0 7.47 ± 0.01 ∆mag 0 Ginski et al.(2016) HD 33564 Calar Alto / AstraLux i0 4.6 ∆mag 1 Ginski et al.(2012) HD 41004 B Hipparcos Hp 8.785 ± 0.014∗ ∆mag 2 Hipparcos and Tycho Catalogues HD 77065 Gemini North / NIRI Ks 6.638 ± 0.020 ∆mag 0 This paper HD 87646 A Hipparcos Hp 8.203 ± 0.002 ∆mag 2 Hipparcos and Tycho Catalogues HD 89744 AEOS I 5.2 mag 2 Roberts et al.(2011) a UKIRT / UFTI H 4.53 mag 0 Mugrauer et al.(2004) HD 104985 Calar Alto / AstraLux i0 8.302 ± 0.145 ∆mag 1 Ginski et al.(2012) HD 114762 Keck / NIRC2 K 5.888 ± 0.017 ∆mag 0 Patience et al.(2002) HD 134113 WIYN / NESSI z0 7.6 ∆mag 0 This paper HD 156279 Calar Alto /AstraLux i0 7.65 ± 0.03 ∆mag 0 Ginski et al.(2016) HD 160508 WIYN / NESSI z0 7.4 ∆mag 0 This paper HD 162020 VLT / NACO L0 6.539 ± 0.024 ∆mag 0 This paper HD 168443 VLT / SPHERE H 5.325 ± 0.016 ∆mag 0 Moutou et al.(2017) VLT / SPHERE Ks 5.211 ± 0.015 ∆mag 0 Moutou et al.(2017) HD 178911 B Hipparcos Hp 6.835 ± 0.013∗ ∆mag 2 Hipparcos and Tycho Catalogues HD 180314 Calar Alto /AstraLux i0 6.14 ± 0.05 ∆mag 0 Ginski et al.(2016) HD 203949 VLT / SPHERE H 3.107 ± 0.200 ∆mag 0 Moutou et al.(2017) VLT / SPHERE Ks 2.994 ± 0.232 ∆mag 0 Moutou et al.(2017) KELT-1 Keck / NIRC2 K0 9.437 ± 0.019 ∆mag 2 Siverd et al.(2012) b Kepler-13 A Palomar HALE / PHARO K0 9.958 ∆mag 0 Adams et al.(2012) WASP-14 Keck / NIRC2 K 8.621 ± 0.019 ∆mag 1 Ngo et al.(2015) WASP-18 VLT / NACO L0 8.131 ± 0.027 ∆mag 0 This paper XO-3 Calar Alto / AstraLux z0 9.798 ∆mag 0 Bergfors et al.(2013) Notes. Curve flags: 0 is contrast curve specific to the observations of the target; 1 is average limits of observed sample (or subset); 2 is typical sensitivity of instrument for the observational set up used (see text for more detail). a we used the typical performance curves given in Turner et al.(2006) for the same observational set up. b as no limits are provided in Siverd et al.(2012), we assumed a similar performance as in Ngo et al.(2015) which used a comparable observing strategy on the same instrument, and used the average detection level from that paper. ∗ magnitude of the primary (see text). limits for our six observed targets (see Section 3.2), we are able to with a 0 in the “curve flag” column in Table7. A number of sur- define a detection probability map which is presented and described veys only provide average detection limits for the observed sample in Section 5.3. (Ginski et al. 2012; Bergfors et al. 2013; Ngo et al. 2015), which we flag as 1. Finally, we considered the typical sensitivities achieved by specific facilities and instruments when no detection limits were 5.1 Imaging contrast curves available. These curves are flagged with a 2 in Table7, and are The data used to derive detection limits for all targets with existing detailed below. imaging data (new or from the literature) are summarised in Ta- Three of our targets have past observations with the AEOS ble7. Sensitivity limits were found to be available for a total of 29 telescope presented in the survey by Roberts et al.(2011). As the objects, including our 6 observed targets. When multiple sets of ob- authors do not provide detection limits for their observations, we servations were found, we chose the best limits available. For a few used the typical performance curves for the AEOS telescope given targets, the deepest contrast curves found only covered a limited in Turner et al.(2006) for the same observation set up as that de- range of separations. In those cases, we also consider the shallower scribed in Roberts et al.(2011), and consider that they are represen- detection limits and keep the best value available at any given pro- tative of those data. Similarly, Siverd et al.(2012) acquired NIRC2 jected separation. images of KELT-1 with Keck but do not present their achieved We used the contrast curves presented in Section 3.2 for the 6 sensitivities. We therefore assumed a similar performance to that targets observed as part of this survey. Most targets with archival achieved in Ngo et al.(2015) for comparable NIRC2 observations observations have contrast curves provided in the literature that are and used the average detection limits from that work for this target. specific to the best set of observations for each target. We flag those Pinfield et al.(2006) searched for companions to 70 Vir out to

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 15

3000using 2MASS. As no detection limits are available in that paper, we generated a 2MASS contrast curve based on the typical resolving and completeness limits of the 2MASS survey. According to the 2MASS documentation (Skrutskie et al. 2006), close dou- bles with separations < 500are not reliably resolved by 2MASS and stellar PSFs can contaminate neighbour sources up to 1000. The J- band completeness limit is given at 16.0 mag. We therefore start our 2MASS contrast curve at 500with ∆J = 0 (equal mass binary at the resolving limit). We then use the completeness limit J = 16.0 from 1000out to 30000, the radial search limit from Pinfield et al.(2006), with a linear increase in ∆J between 500−1000. Finally, three targets in our sample were found to be Hipparcos-Tycho binaries, namely, HD 41004 B, HD 87646 A and HD 178911 B. For those three targets, we used the typical sensitivity to binaries in the Hipparcos catalogue based on the plot of separation against ∆Hp of all Hipparcos binaries found in the ESA documentation (ESA 1997). We extend the separation range out to 3000, the given widest separation of identified Hipparcos-Tycho binaries. For two of these systems, our sample targets correspond to the fainter, lower-mass component of the binary system, which were detected as companions to the brighter primaries. We thus considered the magnitude of the primary of these systems to derive detection limits around the primary binary component. Most of the obtained detection limits are in units of magnitude Figure 10. Detection limits for all targets in our sample with published or difference, ∆mag, while a few are provided in magnitudes and one new direct imaging observations in terms of secondary mass (top) and sys- in flux ratio. These are indicated in the “Limits units” column of Ta- tem mass ratio (bottom). Limits were derived using the data listed in Table ble7. All sensitivity limits are given as a function of angular sepa- 7 and following the approach described in the text. The black stars indicate ration. For all limits that were not in units of magnitudes we started the positions of confirmed companions to the stars with imaging limits and by converting the contrast curves into apparent magnitudes using the open circles correspond to direct imaging candidate companions. the photometry of our targets in the considered filters and given in Table7. Using the distances from Table2, we then converted all magnitude limits into absolute magnitudes and the angular sep- ∆G = 6 were recovered at separations of ∼300. Based on their re- arations into physical projected separations, in AU. Adopting the sults, we define our Gaia DR2 sensitivity limits to start at 100and ages from Table2 for our targets, the obtained absolute magnitude ∆G = 1, with a linear decrease to ∆G = 6 from 100−300. We then curves were then interpolated into the BT-Settl evolutionary models adopt a linear decrease out to 500from ∆G = 6 to G = 21, the Gaia by Allard et al.(2012) to derive corresponding minimum detectable faint limit, and use that limiting magnitude at wider separations, companion masses. The BT-Settl models provide isochrones for nu- out to projected separations corresponding to 104 AU. merous photometric systems. We were therefore able to use models Figure 11 shows the obtained sensitivity limits in terms of ap- corresponding to the specific facilities and filters considered and in- parent G magnitude and mass ratio for all objects in our sample. fer mass limits for each target. Finally, we used the stellar masses We plot on the left panels the limits for the targets without a known listed in Table2 for our sample to convert the obtained mass limits companion and on the right panels the limits for all confirmed mul- into mass ratios q (using the masses of the binary primaries for the tiple systems, with the positions the known companions. Magni- two Hipparcos systems mentioned above). For the few targets with tude limits were converted into corresponding mass ratio curves multiple entries in Table7, we considered the lowest mass ratio adopting the properties of our targets listed in Table2 and follow- value in the overlapping regions of separations in order to define a ing the approach described in the previous section with BT-Settl unique sensitivity curve for each object. The final mass and mass isochrones specific to the Gaia filter system. ratio curves for each target with direct imaging data are shown in While Gaia is essentially complete in the range G∼12−17 Figure 10, together with the positions of all confirmed (black stars) mag, the catalogue has an ill-defined faint magnitude limit which and candidate (open circles) companions around these objects. depends on celestial position (Gaia Collaboration et al. 2018). In addition, the number of sources with a full 5-parameter astrometric solution (position, parallax and proper motion) decreases towards 5.2 Gaia detection limits the faint end, where a larger fraction of sources only have positional Since all objects in our sample are found in the Gaia DR2 cata- measurements available, as discussed in the assessment of the Gaia logue, we are able to derive Gaia detection limits for each of our DR2 astrometric performance by Lindegren et al.(2018). We must targets. Gaia DR2 is found to be complete between G = 12 and take into account the catalogue completeness in our detection lim- G = 17, with a limiting magnitude of G ∼ 21 and a bright limit of its to account for companions that are missed by Gaia, but also for G ∼ 3 (Gaia Collaboration et al. 2018). Ziegler et al.(2018) inves- those like τ Gem B and HAT-P-20 B that only have a 2-parameter tigated the recoverability of close binaries in Gaia DR2 looking for solution and which we were not able to identify as Gaia compan- known binaries from the Robo-AO Kepler survey (Law et al. 2014; ions in our analysis in Section 4.2. Arenou et al.(2018) report the Baranec et al. 2016; Ziegler et al. 2017) in the Gaia DR2 catalogue. completeness of the Gaia DR2 catalogue as a function of G-band They found that near equal-brightness binaries (∆G < 1) were con- magnitude in their catalogue validation work. The provided com- sistently retrieved from separations of 100and that binaries down to pleteness level for sources with full astrometric solutions decreases

MNRAS 000,1–30 (2019) 16 C. Fontanive et al.

Figure 11. Gaia detection limits for all targets without a known companion (left panels) and those that are confirmed binaries or higher order multiples (right panels). The top panels show the Gaia G-band detection limits as defined in the text. Bottoms panels correspond to the same magnitude sensitivities converted into mass ratios using the BT-Settl models (Allard et al. 2012) and the stellar properties from Table2. On the right panels, we also show the positions of the confirmed companions to our targets. Companions detected in Gaia are marked with stars. The two known companions present in Gaia DR2 but with only a 2-parameter astrometric solution are shown by triangles. Binary companions not retrieved in Gaia are indicated with filled circles. from ∼99% at G < 17 to ∼80% at G = 20, before sharply dropping lacking a 5-parameter solution, and that these two sources are not to 0% as G approaches 21 mag (see figure A.1. in the appendix of representative of the completeness of Gaia DR2. Arenou et al. 2018). We thus use the completeness levels provided For all targets in our sample, we converted the completeness in that paper to account for these effects. curve obtained from Arenou et al.(2018) into mass ratios as we In Figure 12 we show an example of Gaia sensitivity curves did for our example target in the bottom panel of Figure 12. We for a primary of G = 8 with a parallax of 20 mas, corresponding to are thus able to assign a completeness factor to each mass ratio a mass of 1 M at 3 Gyr, representative of the targets in our sample. value for every detection limit presented in Figure 11. Instead of In the top panel, we show the detection limits for such an object in traditional sensitivities, where anything above the mass ratio curves terms of apparent magnitude, defined as described above. The hor- is considered as detectable and anything below is not retrievable, izontal dashed lines indicate the G magnitudes associated with var- we now associate every point in the separation-mass ratio space to a ious completeness levels using the information from Arenou et al. detection probability, given by the completeness level at any given (2018), down to the faint magnitude limit of Gaia DR2 at G = 21 mass ratio value. The part of the parameter space below the final (grey line). The bottom panel shows the same contrast curve, con- limits remains at the zero detection probability level regardless of verted into mass ratios using the BT-Settl models and adopting an the associated completeness value. We will use these probabilities age of 3 Gyr. Since we assumed a primary mass of 1 M , the plot in in the next section to define a 2-dimensional detection probability the bottom panel is also representative of the corresponding mass map for our sample. limits in units of Solar masses. Figure 12 clearly demonstrates that when working in mass ratio space the range over which Gaia DR2 is not complete for sources with 5-parameter solutions (<99% com- 5.3 Detection probability map pleteness, below the red dashed lines) is significantly reduced relative to the span of the same incompleteness levels in magnitude We combined all sensitivity curves obtained in Section 5.1 from space, going from G=17−21 to q=0.16−0.09 M . This implies that imaging data and in Section 5.2 from Gaia DR2 to define a single in addition to the targets too faint for Gaia (G > 21, below the detection probability map for our survey, as was done in Fontanive 0% completeness grey dashed lines in Figure 12) only the lowest- et al.(2018). For targets with both Gaia and imaging limits, we mass companions have a high chance to of being missed due to started by combining the two sets of contrast curves. Following the survey incompleteness. We note however that the two known com- approach described in Section 5.1, we considered the best value panions from our sample that only have a 2-parameter astrometric available (lowest q value) in the separation ranges where the Gaia solution in Gaia DR2 (τ Gem B and HAT-P-20 B) have relatively and imaging limits overlapped, keeping track of the ranges over bright G-band magnitudes of 9.42 mag and 12.80 mag, respectively. which the final curves corresponded to the Gaia limits. This al- Their apparent magnitudes fall into a >99% completeness level of lowed us to define a unique sensitivity curve for each object. sources with full astrometric solutions according to Arenou et al. The mass ratio limits for each target in the sample were then (2018). We conclude that it was statistically unlikely to have two placed on a grid of separations and mass ratios with a resolution of bright companions in our sample present in the DR2 catalogue but 0.002 in q and steps of 0.01 in log(ρ). For every cell in the grid,

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 17

Figure 13. Detection probability map for our sample using the mass ratio sensitivities for our targets from Sections 5.1 and 5.2, including the completeness of Gaia DR2 (see Section 5.2). Black contours denote the 0%, 50% and 95% and 99% completeness regions for the full survey. Red stars show the positions of all conﬁrmed companions and yellow circles indicate the positions of candidate companions (see Tables4 and5).

are not detected by Gaia and are more likely to lack a second direct imaging epoch, necessary to be astrometrically confirmed. Interest- ingly, no companion was found at projected separations <20 AU, Figure 12. Completeness of Gaia DR2 compared to the Gaia detection despite reaching a completeness level up to 50%, while a number of limits for a representative target of 1 M at 3 Gyr, with a parallax of 20 companions (confirmed and candidates) were retrieved at the same mas and a G-band magnitude of 8 mag. The top panel shows the contrast detection probability level at wider separations and low mass ratios. curve in terms of apparent magnitude of the secondary and the bottom panel displays the corresponding sensitivity in terms of mass ratio, computed in the same way as the Gaia detection limits for our targets in Figure 12. The coloured dashed lines represent the completeness levels of Gaia DR2 taken 6 STATISTICAL ANALYSIS from Arenou et al.(2018) for sources with a 5-parameter solution. We used the statistical tool described in Fontanive et al.(2018) to constrain the multiplicity properties of our sample. Examining the binary statistics of these objects will allow us to investigate the pos- we then identified the number of targets around which a compan- sible role of binarity in the formation or evolution of massive, close- ion of given separation and mass ratio would have been retrieved in brown dwarfs and planets. The code is based on a Markov Chain given the data gathered for this survey. When the considered sep- Monte Carlo (MCMC) sampling method, using the emcee Python aration corresponded to the Gaia limit of a given target, we then package (Foreman-Mackey et al. 2013), and allows us to place ro- scaled this detection by the Gaia completeness level at the associ- bust Bayesian statistical constraints on the binary frequency and ated mass ratio, which we previously computed in Section 5.2. The companion population distributions for the sample gathered in this number obtained for each cell of the grid was then divided by the study. We add a new capability to the tool in order to account for total number of objects in our sample, providing a value between unconfirmed candidates, which we describe below. 0 and 1 representing the average probability that a companion of The statistical approach uses the detection limits of the survey, projected separation ρ and mass ratio q would have been detected in the form of a detection probability map (see Section 5.3, Fig- around our 38 targets given the data considered in this work. ure 13), and the properties of detected companions (total number Figure 13 shows the resulting detection probability map for of detections, separations and mass ratios of identified systems) to our full sample of 38 objects, considering all available imaging data derive posterior distributions of model parameters (binary fraction for the targets in our survey and the Gaia DR2 catalogue. Compan- and parameters describing the shapes of companion distributions) ions inside the 99% completeness region are essentially detectable most compatible with the gathered data. Based on previous stud- around all targets in the sample. We are complete to companions ies of stellar multiplicity in the field (Duquennoy & Mayor 1991; with q > 0.2 at separations >1000 AU around 90% of our sample, Raghavan et al. 2010), we adopt a lognormal distribution in com- and down to q ∼ 0.1 from separations of ∼100 AU around half panion separation ρ (Equation2) and a power-law in mass ratio q of our targets (50% detection probability contour). Confirmed co- (Equation3): moving companions were found to be relatively evenly distributed P(ρ | µ, σ) ∝ exp [− log (ρ) − µ2 / 2σ2] (2) throughout the parameter space (both in separation and mass ratio). 10 In contrast, most candidate companions are concentrated around where µ and σ are the mean and standard deviation of the underly- q ∼ 0.1, which we attribute to the fact that these fainter companions ing normal distribution in log(ρ). The mass ratio distribution ranges

MNRAS 000,1–30 (2019) 18 C. Fontanive et al. from 0 to 1 and is defined by the power-law index γ: ( qγ for γ 0 P(q | γ) ∝ > (3) (1 − q)−γ for γ < 0 so that negative and positive indices produce symmetric distributions about q = 0.5 for the same absolute value of γ. As was done in Fontanive et al.(2018), we truncated the model distributions at ρ = 20−10,000 AU and q = 0.05−1, in order to constrain the binary frequency on those separation and mass ratio ranges. We adopted flat priors for each model parameter, set to unity over the following ranges and to zero elsewhere: 0.5−4 for µ, 0.1−3 for σ, -3−3 for γ and 0−1 for the binary fraction f . In order to also take into account the candidate companions identified around the targets in our sample, we used the probabilities of the candidates being physically bound derived in Section 4.1 and listed in Table5. At each step in the MCMC chain, we drew for each target with a candidate companion a number between 0 and 1, and counted the candidate as a bonafide companion if the drawn value was below the companionship probability. This ensures that each candidate companion is selected in a fraction of MCMC steps that is representative of its probability of being physically associated to the primary target. For hierarchical systems (e.g. 30 Ari, WASP-14), we considered the properties of all detected components in the part of the code that constrains the shape of the separation and mass ratio distributions, accounting for candidate compan- Figure 14. Posterior probability distributions of all model parameters (diag- ions only when they were selected (e.g. 70 Vir, EPIC 219388192, onal) from our MCMC analysis performed on the full sample of 38 objects HD 89744). For systems in which the binary companion is itself a and correlation among all pairs of parameters (triangle plot). Normalised histograms at the ends of rows are marginalised over all other parameters. tight binary (AS 205, HD 178911, Kepler-13), we used the com- Black contour lines in the correlation plots correspond to regions containing bined mass of the binary component, since this total mass would be 68%, 95% and 99% of the posterior. responsible for any dynamical effect on the close-in planet or brown dwarf. For the section of the tool constraining the binary fraction, we considered the number of multiple systems rather than the total the overall multiplicity rate of FGK stars in the field, generally ob- number of companions in order to constrain the multiplicity rate of served to be around 40−50% (Duquennoy & Mayor 1991; Ragha- our sample. van et al. 2010). The peak of the lognormal in separation corresponds to a value of ∼250 AU, also much wider than for the field population (Raghavan et al. 2010). We discuss these features further in Section8 where we provide a detailed assessment of the 7 RESULTS possible sources and implications of these results. 7.1 MCMC analysis for the full sample We ran the MCMC sampler with 2000 walkers taking 5000 steps 7.2 Sample division at 10 days in inner companion period each on our full sample of 38 objects. We found that walkers were expanding from their initial positions to a reasonable sampling of We divided our sample into two subsets, with a cut at 10 days in the parameter space in less than 100 steps, and removed the first the period of the inner planets and brown dwarfs, the commonly- 100 steps of this “burn-in” phase. We found a mean fraction of accepted threshold for hot Jupiters (Wang et al. 2015; Dawson & steps accepted for each walker of 0.44, in good agreement with the Johnson 2018). This allows us to investigate possible differences in rule of thumb acceptance fraction suggested by Foreman-Mackey the binary properties of the stars with companions on orbits com- et al.(2013) between 0.2 and 0.5, and trust the obtained value to be parable to hot Jupiters, and those with planets or brown dwarfs at a reliable indication of convergence. slightly wider separations. The hot Jupiter-like subset includes 12 The full output from our MCMC analysis is presented in Fig- targets, 6 of which are confirmed binaries, with 2 additional tar- ure 14. The best-fit values for the binary parameters of our core gets having at least one high-probability candidate companion. The sample on separations in the range 20−10,000 AU are summarised sample of wider inner companions contains 26 objects, including in Table8. Errors correspond to 68% confidence intervals, es- 10 confirmed multiples and 1 candidate binary. timated using a highest posterior density approach to determine Following the approach described above, we created detection the boundaries of Bayesian credible intervals (see Fontanive et al. probability maps for each subset, considering the available detec- 2018). The highest density region method provides a set of the most tion limits for all targets from each subsample. We then performed probable values of a given parameter. All 4 model parameters were the same statistical analysis as that presented above to constrain found to be well-constrained, converging to sharply-defined peaks the multiplicity rates and binary properties of our samples of ob- in the posterior distributions, with the binary fraction f showing the jects with periods shorter and longer than 10 days, so as to as- broadest posterior distribution. sess whether statistically significant discrepancies are observed be- The obtained posterior distribution for the binary frequency tween the two populations. Figures 15 and 16 show the output of +13.2 of our sample, f = 79.0−14.7%, is found to be much higher than the MCMC sampler for the shorter and longer-period samples, re-

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 19

Figure 15. Same as Figure 14 for the subset of objects with an inner planet or Figure 16. Same as Figure 14 for the subsample of targets with an inner com- brown dwarf on an orbit shorter than 10 days. panion with a period larger than 10 days.

Table 8. Summary of multiplicity properties from our study, with a comparison to ﬁeld stars and hosts to lower-mass planets.

Sample NM2 µ σ γ f Binary sep. Binary q Reference (MJup) (%) (AU) − +0.14 +0.12 − +0.31 +13.2 − − Full sample 38 7 70 2.39−0.15 0.68−0.10 0.52−0.32 79.0−14.7 20 10,000 0.05 1 This work 70 ± 10 50−2000 0.05−1 This work − +0.28 +0.29 +0.38 +8.0 − − < 10 day subset 12 7 70 2.36−0.31 0.76−0.16 0.03−0.40 92.0−19.0 20 10,000 0.05 1 This work − +0.18 +0.15 − +0.55 +14.4 − − > 10 day subset 26 7 70 2.40−0.17 0.63−0.12 0.89−0.64 74.0−15.9 20 10,000 0.05 1 This work FGK ﬁeld stars 454 − 1.70 1.68 ∼0 44 ± 2 overall overall Raghavan et al.(2010) 36 ± 2 20−10,000 overall Scaled in this work 16 ± 1 50−2000 overall Scaled in Ngo et al.(2016)

Friends of HJs 77 < 4* ...... 47 ± 7 50−2000 0.05−1 Ngo et al.(2016)

*5 objects from the Friends of Hot Jupiters survey have masses between 7−12 MJup, all of which are part of our studied sample (see text). spectively. As expected from the smaller sample sizes of the two crepancy in the output posterior distributions. The obtained prob- subsets relative to the full sample, the walkers are slightly more ability density function for the sample with inner companions on +8.0 widely spread throughout the parameter space than in Figure 14, very short orbits (Figure 15) was found to peak at 92.0−19.0% (68% and this effect is amplified for the smaller sample of <10 days com- confidence), consistent with a binary rate of 100% at the 1-σ level. panions (Figure 15). Nevertheless, all 4 model parameters are still In contrast, the subset of wider inner companions (Figure 16) has +14.4 well-constrained within the explored parameter space in both sub- a binary frequency of 74.0−15.9%. We plot these two distributions samples. The best-fit values and corresponding 1-σ intervals are in Figure 17, together with the obtained posterior distribution for given in Table8 for each subset. f for the full sample. The red line shows the corresponding mul- The model parameters describing the companion separation tiplicity rate of field stars based on results from Raghavan et al. distribution (µ and σ) peak at very similar values for the two sub- (2010), scaled to our probed separation range of 20−10,000 AU sets, indicating that no significant difference is found in the binary (see Section8 for details). While all 3 distributions are consistent separation of these two populations. The power-law index γ de- with one another, a clear shift is observed in the peak of the poste- scribing the mass ratio distribution appears to shift to slightly lower riors. In particular, the peak of the binary fraction distribution for values for the sample of longer-period inner companions, which re- the <10 day sample is located outside or at the edge of the 68% flects the generally lower mass ratios of multiple systems found in confidence interval of the other two posterior distributions. The re- that subset. sulting binary fractions are all much higher than the corresponding The binary fraction f , on the other hand, shows a larger dis- value expected for field stars.

MNRAS 000,1–30 (2019) 20 C. Fontanive et al.

Zucker & Mazeh(2002) noted a substantial paucity of high- mass planets with short-period orbits around single stars. In contrast, this feature is not observed around binary star systems, which exhibit a prevalence of short-orbit massive planets (Zucker & Mazeh 2002; Eggenberger et al. 2004; Desidera & Barbieri 2007; Mugrauer et al. 2007). Our results are highly consistent with these observations, suggesting that almost all stars with a >7 MJup companion within ∼1 AU are part of multiple stellar systems. The statistically higher binary occurrence of hosts to massive planets relative to the general field population indicates that stellar companions may play an important role in the existence of the most massive giant planets and brown dwarfs observed on tight orbits, and that a binary companion may be required to explain their presence. The much higher binary fractions we find for our sample compared to field stars, despite the known biases from transit and radial velocity surveys against close binaries, reinforces the idea of a significant correlation between stellar binarity and the existence of the massive inner companions studied here. While the nature and magnitude of this role are not clear and cannot be established based on this study Figure 17. Posterior probability distributions obtained from our MCMC alone, a number of possibilities have been formulated and explored analysis for the binary frequency of our full sample of 38 objects (solid in the literature to explain the possible influence of binary com- black line), the subset of 12 objects with inner companions on orbits shorter panions on giant planet formation and evolution. We discuss these than 10 days (dashed blue line) and the subsample of 26 systems with theories in Section 8.3. a wider inner companion (dotted grey line). Binary frequencies are con- A caveat of this analysis is that Raghavan et al.(2010) studied strained over the separation range 20−10,000 AU. The vertical lines show stars in a volume limited to 25 pc in distance. In order to compare the positions of the most likely value for each distribution and the corre- our results to the overall field population, we extrapolated the mea- sponding values are indicated above. The ranges of the horizontal lines correspond to the 68% intervals of highest probability. The red line and shaded surements from Raghavan et al.(2010) out to distances of 500 pc. region show the multiplicity fraction of field stars from Raghavan et al. While the distributions found by Raghavan et al.(2010) are valid (2010) which we scaled to the same separation range. for 0.5−1.5 M stars, our sample contains two targets from young star-forming regions (one confirmed binary and one not known to have any companion), as well as a number of giants and subgiants. These populations may have different binary statistics than the main 8 ANALYSIS AND DISCUSSION sequence solar-type stars probed by Raghavan et al.(2010) and the 8.1 Comparison with field stars assumptions made in our analysis may not be entirely valid. The field study by Raghavan et al.(2010) was also heavily biased to- 8.1.1 Multiplicity fraction wards G stars, while our sample contains a number of more massive Raghavan et al.(2010) provided a comprehensive assessment of A and F stars, which are expected to have a higher binary fraction, the multiplicity properties of Solar-type stars, searching for com- as well as some M-dwarfs, expected to have a lower binary fre- panions to 454 F6−K3 primaries in the field. Taking into account quency. The mass dependence of stellar binarity could therefore be the completeness limits of their survey, the authors found that about another factor affecting our results. The field stars sample may also 56 ± 2% of stars are single, for an overall multiplicity fraction of be contaminated with planet hosts, and it must be pointed out that 44 ± 2%, in good agreement with previous results from Duquennoy the results presented above are not a comparison between planet- & Mayor(1991). Our binary fractions derived in Section7 were free stars and planetary hosts, but rather an assessment of planet limited to separations in the range 20−10,000 AU. We must there- hosts multiplicity properties relative to the general stellar popula- fore restrict the overall binary rate from Raghavan et al.(2010) to tion. That being said, the extremely high binary fraction derived for this separation range in order to compare our findings to the gen- our studied sample is still a robust and significant result by itself, eral field population. Taking into account the shape of the distribu- even if the comparison to field stars may not be fully reliable. tions obtained by Raghavan et al.(2010) and excluding all companions from that study with separations outside our considered range, 8.1.2 Mass ratio distribution the fraction of stars found in binaries or higher-order multiples becomes 36 ± 2%. This is more than twice as low as the binary rate Raghavan et al.(2010) found a roughly flat mass ratio distribution obtained for our full sample of 38 objects, with a 3-σ significance for binaries separated by more than 100 days. The value obtained +13.2 ( f = 79.0−14.7%; see Table8). We also find the value for field stars for the power-law index in our full sample indicates a slight pref- to be lower than the binary rates derived for our two separate sub- erence for lower-mass companions, but is fairly consistent with a sets, although these results have a lower significance (∼2.5-σ level) flat distribution (i.e. γ = 0, see Table8). Our subset of >10 day as a result of the smaller number statistics of the individual subsam- inner companions indicates a moderately larger preference towards ples. In Figure 17 we compare the expected fraction of multiples low mass ratio companions for these systems, and the subsample on this separation range to the posterior distributions obtained for of short-period planets was found to exhibit a uniform distribution the full sample studied in this work and to the two subsets defined in wide companion mass ratio. Our probed samples are thus in rea- in Section 7.2. The plot clearly shows that our derived binary rates sonable agreement with the mass ratios observed around multiple are statistically larger than the binary fraction from the overall FGK stars in the field, and we find no evidence for distinct populations stellar population. between our studied targets and the general field population.

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 21

close binary systems. In addition, it may be harder to detect planets via the radial velocity method in the presence of a close, massive stellar companion. The spurious assumption that a planet host is single when it is in fact an unresolved binary will also lead to er- roneous measurements of the planet’s physical properties. A population of massive planets and brown dwarfs in closely-separated binaries could hence exist and be underrepresented or misreported among detected exoplanets. If this is the case, the true multiplicity rate of systems hosting massive, close-in planets or brown dwarfs should then be even higher than what we found here. Nevertheless, our observations are consistent with previous studies. The shortfall of close binaries among planet hosts has indeed been vastly reported in the literature (Roell et al. 2012; Bergfors et al. 2013; Wang et al. 2014a,b; Kraus et al. 2016) and is generally attributed to a hindrance of planet formation in very Figure 18. Separation distributions of wide binary or hierarchical compan- tight binaries. While observational constraints remain sparse, the ions, comparing the output from our MCMC analysis on the full sample current census is that binarity on scales comparable to the Solar of 38 objects (solid black line) and two subsets (dashed blue line and grey system (.50−100 AU) has the potential to affect planet formation dotted line), to the field population from Raghavan et al.(2010) in red. Our and evolution. However, different conclusions have been reached obtained density functions show narrower distributions, peaking at larger on the theoretical side, and it is not clear whether these processes separations than for field stars. are altered, inhibited or facilitated by the presence of a binary companion and on what separations these effects may take place. This is further discussed in Section 8.3. 8.1.3 Separation distribution Another possible explanation for the depletion with tight bi- In contrast, we found larger and more significant disparities in binary systems among our sample is that the inner planetary and nary companion separation between the distributions obtained in brown dwarf companions to our targets may have formed at much Section7 and the expected distribution from FGK field stars, as wider separations than their current locations, at radii overlapping shown in Table8. Raghavan et al.(2010) reported a lognormal with the missing population of binary companions. Such massive planets are indeed expected to form outside at least a few AU for distribution in companion separation peaking at 1.70 in log10(a), with a Gaussian width of 1.68, corresponding to a broad peak formation by CA (Mordasini et al. 2012), and more likely several around 50 AU. This is significantly smaller than our derived value tens of AU for GI, in regions of the circumstellar discs that are mas- +0.14 sive and cool enough to form massive giant planets (Rafikov 2005). of µ = 2.39−0.15 for our full sample, with a mean located at ∼250 AU. We also found a much narrower separation distribution, with a If inner companions have formed at such separations, additional, +0.12 massive binary companions should not exist within a few tens to Gaussian width of σ = 0.68−0.10. The results obtained for our two subsets are in good agreement with each other and with the full hundreds of AU around these systems, which would be reconcil- sample (Table8). Figure 18 shows the constraints obtained from able with our observations. However, as this trend is also observed our statistical analysis on the separation distribution of the multi- around systems hosting low-mass planets, for which a formation at ple systems in our core sample and defined subsets. The red dis- very wide separations is not required, this may not be the primary tribution represents the results obtained by Raghavan et al.(2010) phenomenon responsible for this feature. for solar-type stars in the field, clearly demonstrating the preference for wider binaries among our targets and the strong deficit of closely-separated systems in our studied sample. 8.2 Binarity as a function of planet properties While we restricted ourselves to a 20−10,000 AU separation 8.2.1 Binary frequency versus inner companion period range to constrain the binary frequency f , the parameters describing the separation distribution (µ and σ) were explored over a broad In Section 7.2, we divided our sample into two subsets in order to parameter space. The MCMC walkers would have been able to con- investigate possible differences in the demographics of stars host- verge to a distribution peaking near or even inside 20 AU had it ing planetary or brown dwarf companions within and beyond or- been compatible with the observed data. As we noted in Section bital periods of 10 days. While we found no evidence for distinct 5.3, the lack of tight binaries is unlikely to be the sole result of binary mass ratio or separation distributions between these two observational biases and limiting sensitivities. Indeed, we are sen- populations, our statistical analysis revealed a possibly larger bi- sitive to binary companions with separations of 5−20 AU around nary frequency for the subset of shorter-period companions, with a 20−60% of our targets depending on the system mass ratio, and a peak at 92%, compared to 74% for the subsample of more widely- number of companions (confirmed and candidates) are retrieved at separated systems. These results are marginal due to the less strin- the same detection probability level at larger separations and low- gent constraints we were able to place on the individual subsets, as mass ratios (see Figure 13). If the true underlying separation dis- shown by the broader posterior distributions in Figure 17 relative tribution in our sample was comparable to that of field stars, with the one obtained for the full sample. Larger sample sizes will be a broad peak near 50 AU, we would have expected to detect nu- required to confirm this tendency. merous close binaries given the number of wide multiple systems This theory is nonetheless supported by the similar trend seen already present in the sample. We thus consider that this observed for hosts to lower-mass planets. Surveys searching for wide com- feature is real and not due to observational limitations. panions to radial velocity planets from sub-Jupiter masses up to a This deficiency of tight binaries could however be due to se- few MJup and out to 5 AU found that less than ∼25% of these sys- lection effects in exoplanet surveys, which are often biased against tems were part of binaries or multiple systems (e.g. Raghavan et al.

MNRAS 000,1–30 (2019) 22 C. Fontanive et al.

2006; Ginski et al. 2012), although these surveys may be biased or in Figure5. Given that their mass measurements are lower lim- incomplete. In contrast, studies of slightly shorter-period transiting its derived from radial velocity information, their true masses are planets (typically <100 days) observed rather higher binary frac- most certainly even higher. Assuming a uniform distribution of in- tions, generally around ∼50%, for planets of comparable masses clinations, we may calculate the minimum value for the projected (Adams et al. 2012, 2013; Ngo et al. 2016). Furthermore, Tokovinin mass that corresponds to a true substellar mass M2 < 70 MJup at et al.(2006) found that 96% of spectroscopic binaries with periods a given confidence level. This translates to M2 sin i < 34 MJup for <3 days have a third component, compared to only 34% of systems a 68% confidence of a true mass M2 below the hydrogen-burning with periods longer than 12 days, albeit some selection biases may limit. These four targets thus have a large chance of being stellar, affect these results to a limited extent. and are therefore likely to have formed as tight stellar binary sys- The marginal difference in binary occurrence observed in tems, rather than brown dwarf companions forming in a circumstel- this work between very short-period transiting planets and brown lar disc around the host star. None of these systems were found to dwarfs, and the somewhat wider population of radial velocity com- have a wide binary companion. Excluding these systems from our panions, thus indicates that this trend, if real, may also hold for survey to focus on substellar companions only would hence have the very massive inner companions studied here. This trend could resulted in an even higher binary fraction for the rest of our sam- suggest that binarity greatly helps the formation or migration of ple. This further reinforces the idea that the most massive planets massive giant planets and brown dwarfs observed within ∼1 AU, and brown dwarfs forming in discs and detected within ∼1 AU re- and essentially becomes necessary for these companions to reach quire a wide stellar companion to form or evolve to their observed orbital periods shorter than 10 days. orbital configurations.

8.2.2 Binary frequency versus inner companion mass 8.3 Implications for formation and evolution processes In the final paper of the Friends of hot Jupiters campaign (Knutson Our results demonstrate a very robust correlation between binary et al. 2014; Piskorz et al. 2015; Ngo et al. 2015), Ngo et al.(2016) occurrence rate and the sporadic population of close-in massive found that 47±7% of hot Jupiter systems have a stellar compan- giant planets and brown dwarf desert inhabitants. Whatever the ion between 50 and 2000 AU, a binary rate 3 times higher that for underlying processes, this concurrence implies that wide binaries field stars in this separation range. The authors concluded that bi- must have an influence on the observed population of short-period nary companions on these separations facilitate planet formation planetary and brown dwarf companions, which could occur at the or help the inward migration of giant planets. Our study probed stage of formation or during later evolution. higher-mass planets than those considered in that survey, allowing Zucker & Mazeh(2002) were the first to raise the possibility us to examine trends in stellar multiplicity as a function of planet that planets in binaries may have a different mass-period distribu- mass, including inside the brown dwarf regime. The Friends of hot tion, a trend subsequently confirmed by Eggenberger et al.(2004), Jupiters survey looked at systems with planet masses mostly lim- Desidera & Barbieri(2007), Mugrauer et al.(2007) and others. Our ited to 4 MJup. Only five objects with more massive companions results are in very good agreement with these studies, suggesting were studied in that work, with masses between 7 and 12 MJup, all that the most massive planets observed within ∼1 AU are almost ex- of which are part of our selected targets (3 are confirmed multiples: clusively found in binary systems, and that this feature is amplified HAT-P-20, WASP-14 and WASP-18; 2 are apparently single: HAT- as planets or brown dwarfs reach shorter periods. Desidera & Bar- P-2 and XO-3). The binary fraction derived here for more massive bieri(2007) concluded that the presence of a stellar companion on objects was estimated for separations between 20 and 10,000 AU. separations <100−300 AU may be able to modify the formation or The corresponding binary rate restricted to the 50−2000 AU sepa- evolution of giant planets. Eggenberger et al.(2004) also found that ration range becomes 70±10% for our core sample, 1.5 times larger massive planets in binary systems with periods shorter than 40 days than the value from Ngo et al.(2016) at the 2.3- σ level. This frac- have very low eccentricities, suggesting that these planets likely tion is 4 times larger for field stars on these separations (see Ngo underwent some form of migration, possibly induced or driven by et al. 2016), with a 5-σ significance. These results are summarised outer binary companions, to evolve to their current orbits. Duchêne in Table8. (2010) investigated these observational trends and concluded that Our results for the shorter-period subset are less significant binarity does not affect the formation and growth of planetesimals due to the looser constraints we were able to place on the smaller- (see also Batygin et al. 2011 and Rafikov 2013). Duchêne(2010) sized subsample. We thus only compare previous studies with the proposed that planet formation in binaries tighter than ∼100 AU oc- binary fraction estimated for the full sample, and keep in mind that curs at a similar rate but through different mechanisms than around hosts to shorter-period planets may have an enhanced binary rate, wider binaries and single stars, possibly explaining the observed as discussed previously. Our findings suggest that the trends char- preponderance of very massive, close-in planets found in binaries acterised by Ngo et al.(2016) for hosts to hot Jupiters are also ob- but rarely seen around isolated stars. served and even strengthened for the highest-mass close-in planets Simulations by Kley(2001) showed that perturbations from a and brown dwarfs. These results are in excellent agreement with secondary star may alter the formation and evolution of a planet, early observations by Zucker & Mazeh(2002) and Eggenberger in particular by enhancing the mass accretion and orbital migra- et al.(2004), who determined that the most massive planets on or- tion rates. This could explain why the most massive short-period bits of a few days are consistently found in binary systems, sug- planets are found in multiple systems, the presence of stellar com- gesting that this planetary population does not exist around single panions enabling massive planets to achieve smaller orbital separa- stars. tions than the corresponding limit for planets orbiting single stars We note that four targets in our sample have notably high mass (Eggenberger et al. 2004). Jensen & Akeson(2003) found that the estimates (40−60 MJup) relative to the rest of our sample, namely distribution of disc mass in >200 AU binary systems among T Tauri BD+24 4697, HD 77065, HD 134114 and HD 160508. These ob- stars is not always determined by the stellar masses and may be jects appear somewhat isolated in the period-mass parameter space more asymmetric, with the primary retaining a much more sub-

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 23 stantial disc and the secondary being left with a very low-mass disc. Massive discs around primaries in wide binaries could thus provide larger reservoirs of material for planet formation, which is thought to be favourable to the formation of higher-mass planets, as discussed in Mordasini et al.(2012). The shorter lifetime of circumstellar discs in tight binaries (e.g. Kraus et al. 2012) ar- gues for a formation via gravitational collapse of the circumstellar disc (thousand year timescale) rather than through core accretion, which requires 1−10 million years. A favoured formation by gravitational disc instability is further supported by the very high masses of the giant planets or brown dwarfs considered here. Furthermore, theoretical work by Boss(2006) suggested that a close stellar companion could rapidly induce gravitational perturbations and trigger the instabilities needed for gravitational fragmentation to proceed, even if the disc is not initially unstable to its own gravity. However, simulations by Forgan & Rice(2009) indicate that, rather than pro- Figure 19. Minimum companion mass ratios necessary to excite Kozai- moting fragmentation, perturbations from an outer companion are Lidov oscillations for a 15 MJup planet with an initial semi-major axis of 1 more likely to make the disc more stable. AU, 5 AU, 10 AU and 20 AU around a 1 M star, as a function of wide com- The brown dwarf desert is thought to be a natural feature aris- panion separation. Companions lying to the left of each line are close and ing from formation around single stars, where massive objects with massive enough to induce Kozai-Lidov oscillations and overcome the peri- brown dwarf masses can only form at wide separations and can center precession of inner planets at 1, 5, 10 and 20 AU (see text). We over- plot the positions of the confirmed (black symbols) and candidate (white be challenging to bring inwards through disc migration alone. By symbols) binaries in our Kozai-consistent subsample. modifying the circumstellar disc environment, allowing for different conditions facilitating in-situ formation, and/or by triggering migration processes, the presence of a binary companion could help scenario. In Section 7.2, we showed that this subsample may have populate the low-mass end of the brown dwarf desert and explain a marginally higher binary fraction than the already very high mul- the puzzling existence of the scarce population of very close-in tiplicity rate observed for our full sample. This is thus compatible brown dwarf desert inhabitants. with the idea that the inner companions from this subsets could It is worth noting that over half of detected binary compan- have been driven to their current orbital configurations through ions in our study have projected separations larger than 200 AU. Kozai-Lidov oscillations under the effect of wide companions. We We therefore argue that wider binaries must also be able to impact, also note that 4 of our 5 targets also studied in the Friends of hot almost to the same degree, the formation and/or evolution of these Jupiters campaign have high obliquities (placed in the “misaligned” systems. The processes described above must therefore also be pos- sample in Knutson et al. 2014), a feature often associated with the sible from wider separations in order to account for the existence of Kozai-Lidov mechanism. the planets and brown dwarfs probed in this work. An easy way to Unfortunately, full orbital parametrisation, including inclina- facilitate this is to form the inner companions at significantly larger tion measurements, is not possible for wide, directly-imaged bi- orbital distances (tens of AU), increasing the initial gravitational naries. Nevertheless, we can determine if the observed wide com- influence of the outer companion. As mentioned previously, this panions could be responsible for a Kozai-Lidov scattering of the theory could tentatively explain the shortfall of binaries with sep- inner planets and brown dwarfs based on their masses and orbital arations <50−100 AU, which would then not be expected around distances. This is done by estimating the minimum companion such systems. mass required to excite Kozai-Lidov oscillations on a timescale shorter than general pericenter precession, as was done in Ngo et al.(2016). We adopted a primary stellar mass of 1 M and a 8.4 Scattering and migration via the Kozai-Lidov mechanism mass of 15 MJup for the inner companion, close to the median One way binary companions could assist the migration of these of our Kozai-consistent sample. Equating equations (1) and (23) systems is through the Kozai-Lidov mechanism (Kozai 1962; Li- from Fabrycky & Tremaine(2007) for the Kozai-Lidov oscillation dov 1962). In this alternative scenario to produce hot Jupiters, an timescale and pericenter precession due to general relativity, re- outer binary companion on an inclined orbit relative to the orbital spectively, we computed in Figure 19 the minimum masses and plane of the planet triggers periodic oscillations of the planet’s ec- separations necessary to migrate a 15 MJup companion with initial centricity and inclination (Fabrycky & Tremaine 2007; Wu et al. orbital separations of 1 AU, 5 AU, 10 AU and 20 AU through this 2007; Dong et al. 2014; Petrovich 2015). Combined with the ef- scenario. We assumed initially circular orbits for the inner compan- fects of tidal dissipation, these secular interactions can result in a ions and eccentricities of 0.5 for the outer companions, based on the very short orbit for the inner companion (Rice et al. 2015), pos- roughly uniform eccentricity distribution between 0 and 1 found by sibly with a high spin-orbit misalignment with its host star. The Raghavan et al.(2010) for wide field binaries. We found that almost amplitude of these interactions mostly depends on the initial mu- all detected binary companions could explain the presence of the tual inclination between the inner and outer companions (Fabrycky inner companions in this subset via the Kozai-Lidov mechanism, & Tremaine 2007), allowing Kozai-Lidov cycles to be induced by assuming they formed at separations larger than at least 1−10 AU. very distant perturbers. These hierarchical systems are hence compatible with a mi- As noted in Section2, the subset of targets with a planet or gration of the inner companions through the Kozai-Lidov scenario brown dwarf within a 10 day orbit corresponds to the systems in based on this simple analysis. However, the subset of objects in- our sample with a tidal circularisation timescale shorter than ∼15 consistent the Kozai-Lidov mechanism, based on our tidal circular- Gyr, which we consider to be fully consistent with the Kozai-Lidov isation timescale argument (see Section2), was also found to have

MNRAS 000,1–30 (2019) 24 C. Fontanive et al. a particularly large binary frequency. This suggests that these sys- larger binary rate than systems with longer-period inner compan- tems do not primarily migrate via Kozai-Lidov oscillations. This is ions, consistent with a 100% multiplicity fraction at the 1-σ level. in good agreement with the theoretical study by Naoz et al.(2012) If confirmed, this trend could suggest that the influence of binarity and observational constraints placed by Ngo et al.(2016), which on the formation/evolution of the most massive planetary compan- concurred that only 20 to 30% of all hot Jupiters can be explained ions is enhanced for shorter-period planets, and may even become by the Kozai-Lidov migration process. Kozai-Lidov oscillations are a requirement for the very closest planets and brown dwarfs. therefore unlikely to be the dominant mechanism driving close-in We conclude that wide binary companions have a crucial in- massive planets to their current locations. fluence on planet formation and/or evolution and may be responsible for the sporadic population of high-mass planets and brown dwarf desert members observed on very tight orbital configurations, which seem to rarely exist around isolated stars. 9 SUMMARY AND CONCLUSIONS We have gathered a compilation of 38 planets or brown dwarfs with masses of at least 7 MJup and orbiting within ∼1 AU from their host ACKNOWLEDGEMENTS stars, with the aim of examining the multiplicity statistics of these We thank our anonymous referee for their insightful comments and systems. We searched for wide binary companions to these objects suggestions. This paper comes from work undertaken during a visit using new direct imaging data, observations reported in the liter- funded by the Scottish Universties Physics Alliance (SUPA) Post- ature and the Gaia DR2 catalogue. A total of 16 confirmed mul- graduate, Postdoctoral and Early Career Researcher Short-Term tiple systems were found, and another 3 targets have at least one Visits funding. This work benefited from the Exoplanet Summer high-probability candidate companion. We report here the discov- Program in the Other Worlds Laboratory (OWL) at the University ery in Gaia DR2 of a new mid-K tertiary component comoving of California, Santa Cruz, a program funded by the Heising-Simons with the WASP-14 binary system, and present an independent con- Foundation. KM acknowledges funding by the Science and Tech- firmation of the wide M7.5 companion to WASP-18. We used a ro- nology Foundation of Portugal (FCT), grants No. IF/00194/2015 bust MCMC statistical approach to constrain the binary properties and PTDC/FIS-AST/28731/2017. This work has made use of data of our sample, correcting for observational biases and incomplete- from the European Space Agency (ESA) mission Gaia (https: ness. Our main results are summarised below. //www.cosmos.esa.int/gaia), processed by the Gaia Data Pro- 1. A very high binary fraction. Our analysis revealed a very cessing and Analysis Consortium (DPAC, https://www.cosmos. large binary frequency of f = 79.0+13.2% for these outer com- −14.7 esa.int/web/gaia/dpac/consortium). This study is based on panions on separations between 20 and 10,000 AU, which is more observations obtained at the Gemini Observatory, which is operated than twice as high as for field stars on the same separation range, by the Association of Universities for Research in Astronomy, Inc., with a 3-σ significance. These results demonstrate that wide bi- under a cooperative agreement with the NSF on behalf of the Gem- nary companions greatly influence the formation or evolution of ini partnership: the National Science Foundation (United States), these close-in massive planets and brown dwarfs. The presence of National Research Council (Canada), CONICYT (Chile), Minis- a binary companion could allow for different natal environments in terio de Ciencia, Tecnología e Innovación Productiva (Argentina), circumstellar discs, enabling in-situ formation at locations where Ministério da Ciência, Tecnologia e Inovação (Brazil), and Korea giant planet formation is not normally possible. Stellar companions Astronomy and Space Science Institute (Republic of Korea). This could also facilitate disc migration beyond the extent normally seen work is based on observations collected at the European Southern around single stars, or could trigger alternative migration processes Observatory under ESO programme 099.C-0728. Some of the ob- through induced secular interactions. servations in the paper made use of the NN-EXPLORE Exoplanet 2. A deficit of close binaries. The output of our statistical anal- and Stellar Speckle Imager (NESSI). NESSI was funded by the ysis also showed a strong preference for wide binaries, with a peak NASA Exoplanet Exploration Program and the NASA Ames Re- around 250 AU, compared to ∼50 AU for the overall field popula- search Center. NESSI was built at the Ames Research Center by tion. The apparent shortfall of <50−100 AU binaries is consistent Steve B. Howell, Nic Scott, Elliott P. Horch, and Emmett Quigley. with previous studies. It is not clear whether this deficiency indi- This research has made use of the NASA Exoplanet Archive, which cates that planet formation is inhibited in tight binaries, that our is operated by the California Institute of Technology, under contract probed planets formed near these separations, or if it is the result with the National Aeronautics and Space Administration under the of selection biases in exoplanet surveys. Based on these observa- Exoplanet Exploration Program. This research has made use of the tions, we argue that the mechanisms assisting planet formation or Exoplanet Orbit Database, and the Exoplanet Data Explorer at ex- evolution in multiple star systems must be associated with widely- oplanets.org, as well as the Extrasolar Planets Encyclopaedia at ex- separated binaries, on distances larger than several hundreds of AU. oplanet.eu. This publication makes use of data products from the However, we did find that the Kozai-Lidov mechanism is unlikely Two Micron All Sky Survey, which is a joint project of the Uni- to be the dominant underlying process. versity of Massachusetts and the Infrared Processing and Analysis 3. A higher binary rate for higher-mass planets. A comparison Center/California Institute of Technology, funded by the National with the population of lower-mass planets suggests that binary oc- Aeronautics and Space Administration and the National Science currence increases with planet mass for these close-in objects. This Foundation. This research has made use of the SIMBAD database, is in good agreement with prior studies that found the most massive operated at CDS, Strasbourg, France. planets to be almost exclusively observed in binary systems, and indicates that the role played by binary companions in the existence of these systems becomes more critical for higher-mass planets. 4. A higher binary rate for shorter-period planets. Dividing REFERENCES our sample into two subsets, we found that stars hosting planets Adams E. R., Ciardi D. R., Dupree A. K., Gautier T. N. I., Kulesa C., Mc- or brown dwarfs with orbital periods <10 days have a marginally Carthy D., 2012,AJ, 144, 42

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 25

Adams E. R., Dupree A. K., Kulesa C., McCarthy D., 2013,AJ, 146, 9 Eggenberger A., Udry S., Chauvin G., Beuzit J.-L., Lagrange A.-M., Sé- Aguilera-Gómez C., Ramírez I., Chanamé J., 2018, A&A, 614, A55 gransan D., Mayor M., 2007, A&A, 474, 273 Allard F., Homeier D., Freytag B., Sharp C. M., 2012, in Reylé Eggenberger A., Udry S., Chauvin G., Forveille T., Beuzit J.-L., Lagrange C., Charbonnel C., Schultheis M., eds, EAS Publications Series A.-M., Mayor M., 2011, in Sozzetti A., Lattanzi M. G., Boss A. P., Vol. 57, EAS Publications Series. pp 3–43 (arXiv:1206.1021), eds, IAU Symposium Vol. 276, The Astrophysics of Planetary Sys- doi:10.1051/eas/1257001 tems: Formation, Structure, and Dynamical Evolution. pp 409–410 Almeida P. V., Gameiro J. F., Petrov P. P., Melo C., Santos N. C., Figueira (arXiv:1101.0432), doi:10.1017/S1743921311020564 P., Alencar S. H. P., 2017, A&A, 600, A84 Eggleton P. P., Tokovinin A. A., 2008, MNRAS, 389, 869 Andrews S. M., Williams J. P., 2007, ApJ, 659, 705 Eggleton P. P., Kiseleva L. G., Hut P., 1998, ApJ, 499, 853 Arenou F., et al., 2018, A&A, 616, A17 Eisner J. A., Hillenbrand L. A., White R. J., Akeson R. L., Sargent A. I., Augereau J. C., Papaloizou J. C. B., 2004, A&A, 414, 1153 2005, ApJ, 623, 952 Bailer-Jones C. A. L., Rybizki J., Fouesneau M., Mantelet G., Andrae R., Esteves L. J., De Mooij E. J. W., Jayawardhana R., 2015, ApJ, 804, 150 2018,AJ, 156, 58 Fabrycky D., Tremaine S., 2007, ApJ, 669, 1298 Bakos G. Á., et al., 2011, ApJ, 742, 116 Filippazzo J. C., Rice E. L., Faherty J., Cruz K. L., Van Gordon M. M., Baranec C., Ziegler C., Law N. M., Morton T., Riddle R., Atkinson D., Looper D. L., 2015, ApJ, 810, 158 Schonhut J., Crepp J., 2016,AJ, 152, 18 Fontanive C., Biller B., Bonavita M., Allers K., 2018, MNRAS, 479, 2702 Batygin K., Morbidelli A., Tsiganis K., 2011, A&A, 533, A7 Foreman-Mackey D., Hogg D. W., Lang D., Goodman J., 2013, PASP, 125, Bell K. R., Cassen P. M., Klahr H. H., Henning T., 1997, ApJ, 486, 372 306 Bergfors C., et al., 2013, MNRAS, 428, 182 Forgan D., Rice K., 2009, MNRAS, 400, 2022 Bonfanti A., Ortolani S., Nascimbeni V., 2016, A&A, 585, A5 Forgan D., Rice K., 2011, MNRAS, 417, 1928 Bonomo A. S., et al., 2017, A&A, 602, A107 Gaia Collaboration et al., 2016, A&A, 595, A1 Boss A. P., 1998, ApJ, 503, 923 Gaia Collaboration et al., 2018, A&A, 616, A1 Boss A. P., 2006, ApJ, 641, 1148 Galland F., Lagrange A.-M., Udry S., Chelli A., Pepe F., Beuzit J.-L., Mayor Bowler B. P., Liu M. C., Cushing M. C., 2009, ApJ, 706, 1114 M., 2005, A&A, 444, L21 Galland F., Lagrange A.-M., Udry S., Beuzit J.-L., Pepe F., Mayor M., 2006, Brandner W., et al., 2000,AJ, 120, 950 A&A, 452, 709 Butler R. P., Marcy G. W., Williams E., Hauser H., Shirts P., 1997, ApJ, Ghez A. M., Neugebauer G., Matthews K., 1993,AJ, 106, 2005 474, L115 Ginski C., Mugrauer M., Seeliger M., Eisenbeiss T., 2012, MNRAS, 421, Butler R. P., Marcy G. W., Fischer D. A., Brown T. M., Contos A. R., Ko- 2498 rzennik S. G., Nisenson P., Noyes R. W., 1999, ApJ, 526, 916 Ginski C., et al., 2016, MNRAS, 457, 2173 Butler R. P., et al., 2006, ApJ, 646, 505 Grether D., Lineweaver C. H., 2006, The Astrophysical Journal, 640, 1051 Chauvin G., Lagrange A.-M., Udry S., Fusco T., Galland F., Naef D., Beuzit Guenther E. W., Hartmann M., Esposito M., Hatzes A. P., Cusano F., Gan- J.-L., Mayor M., 2006, A&A, 456, 1165 dolfi D., 2009, A&A, 507, 1659 Chauvin G., et al., 2012, A&A, 542, A41 Guilloteau S., Simon M., Piétu V., Di Folco E., Dutrey A., Prato L., Chapil- Chen X. P., Henning T., van Boekel R., Grady C. A., 2006, A&A, 445, 331 lon E., 2014, A&A, 567, A117 Clarke C. J., Lodato G., 2009, MNRAS, 398, L6 Hall C., Forgan D., Rice K., 2017, MNRAS, 470, 2517 Coker C. T., Gaudi B. S., Pogge R. W., Horch E., 2018, AJ, 155, 27 Hellier C., et al., 2009, Nature, 460, 1098 Csizmadia S., Hellard H., Smith A. M. S., 2019, A&A, 623, A45 Herriot G., et al., 2000, in Wizinowich P. L., ed., Proc. SPIEVol. Curiel S., Cantó J., Georgiev L., Chávez C. E., Poveda A., 2011, A&A, 525, 4007, Adaptive Optical Systems Technology. pp 115–125, A78 doi:10.1117/12.390288 Curtis J. L., Wolfgang A., Wright J. T., Brewer J. M., Johnson J. A., 2013, Hodapp K. W., et al., 2003, PASP, 115, 1388 AJ, 145, 134 Horch E. P., Veillette D. R., Baena Gallé R., Shah S. C., O’Rielly G. V., van Daemgen S., Hormuth F., Brandner W., Bergfors C., Janson M., Hippler S., Altena W. F., 2009,AJ, 137, 5057 Henning T., 2009, A&A, 498, 567 Horch E. P., Howell S. B., Everett M. E., Ciardi D. R., 2012,AJ, 144, 165 Daemgen S., Bonavita M., Jayawardhana R., Lafrenière D., Janson M., Hußmann B., Stolte A., Brandner W., Gennaro M., Liermann A., 2012, 2015, ApJ, 799, 155 A&A, 540, A57 Dawson R. I., Johnson J. A., 2018, Annual Review of Astronomy and As- Irwin J., et al., 2010, ApJ, 718, 1353 trophysics, 56, 175 Jang-Condell H., Mugrauer M., Schmidt T., 2008, ApJ, 683, L191 Desidera S., Barbieri M., 2007, A&A, 462, 345 Jenkins J. S., et al., 2017, MNRAS, 466, 443 Desidera S., et al., 2004, in Beaulieu J., Lecavelier Des Etangs A., Terquem Jensen E. L. N., Akeson R. L., 2003, ApJ, 584, 875 C., eds, Astronomical Society of the Pacific Conference Series Vol.321, Jofré E., Petrucci R., Saffe C., Saker L., de la Villarmois E. A., Chavero C., Extrasolar Planets: Today and Tomorrow. p. 103 Gómez M., Mauas P. J. D., 2015, A&A, 574, A50 Díaz R. F., et al., 2012, A&A, 538, A113 Johansen A., Oishi J. S., Mac Low M.-M., Klahr H., Henning T., Youdin Diolaiti E., Bendinelli O., Bonaccini D., Close L., Currie D., Parmeggiani A., 2007, Nature, 448, 1022 G., 2000, A&AS, 147, 335 Johns-Krull C. M., et al., 2016, ApJ, 826, 206 Dobbs-Dixon I., Lin D. N. C., Mardling R. A., 2004, ApJ, 610, 464 Johnson J. A., et al., 2011, ApJ, 735, 24 Döllinger M. P., Hatzes A. P., Pasquini L., Guenther E. W., Hartmann M., Johnson M. C., Cochran W. D., Albrecht S., Dodson-Robinson S. E., Winn Girardi L., Esposito M., 2007, A&A, 472, 649 J. N., Gullikson K., 2014, ApJ, 790, 30 Dommanget J., Nys O., 2000, A&A, 363, 991 Jones M. I., Jenkins J. S., Rojo P., Melo C. H. F., Bluhm P., 2013, A&A, Dong S., Katz B., Socrates A., 2014, ApJ, 781, L5 556, A78 Dong R., Zhu Z., Fung J., Rafikov R., Chiang E., Wagner K., 2016, ApJ, Jones M. I., Jenkins J. S., Bluhm P., Rojo P., Melo C. H. F., 2014, A&A, 816, L12 566, A113 Duchêne G., 2010, ApJ, 709, L114 Jones J., et al., 2015, ApJ, 813, 58 Duquennoy A., Mayor M., 1991, A&A, 248, 485 Kaib N. A., Raymond S. N., Duncan M., 2013, Nature, 493, 381 ESA ed. 1997, The HIPPARCOS and TYCHO catalogues. Astrometric and Kane S. R., et al., 2011a, ApJ, 733, 28 photometric star catalogues derived from the ESA HIPPARCOS Space Kane S. R., Henry G. W., Dragomir D., Fischer D. A., Howard A. W., Wang Astrometry Mission ESA Special Publication Vol. 1200 X., Wright J. T., 2011b, ApJ, 735, L41 Eggenberger A., Udry S., Mayor M., 2004, A&A, 417, 353 Kane S. R., et al., 2015, ApJ, 815, 32

MNRAS 000,1–30 (2019) 26 C. Fontanive et al.

Kley W., 2001, in Zinnecker H., Mathieu R., eds, IAU Symposium Vol. 200, Piskorz D., Knutson H. A., Ngo H., Muirhead P. S., Batygin K., Crepp J. R., The Formation of Binary Stars. p. 511 Hinkley S., Morton T. D., 2015, ApJ, 814, 148 Knutson H. A., et al., 2014, ApJ, 785, 126 Pollack J. B., Hubickyj O., Bodenheimer P., Lissauer J. J., Podolak M., Kozai Y., 1962,AJ, 67, 591 Greenzweig Y., 1996, Icarus, 124, 62 Kratter K. M., Murray-Clay R. A., Youdin A. N., 2010, ApJ, 710, 1375 Prato L., Greene T. P., Simon M., 2003, ApJ, 584, 853 Kraus A. L., Ireland M. J., Hillenbrand L. A., Martinache F., 2012, ApJ, Rafikov R. R., 2005, ApJ, 621, L69 745, 19 Rafikov R. R., 2013, ApJ, 765, L8 Kraus A. L., Ireland M. J., Huber D., Mann A. W., Dupuy T. J., 2016,AJ, Raghavan D., Henry T. J., Mason B. D., Subasavage J. P., Jao W.-C., 152, 8 Beaulieu T. D., Hambly N. C., 2006, ApJ, 646, 523 Lafrenière D., Jayawardhana R., van Kerkwijk M. H., 2008, ApJ, 689, L153 Raghavan D., et al., 2010, ApJS, 190, 1 Lafrenière D., Jayawardhana R., van Kerkwijk M. H., Brandeker A., Janson Reboussin L., Guilloteau S., Simon M., Grosso N., Wakelam V., Di Folco M., 2014, ApJ, 785, 47 E., Dutrey A., Piétu V., 2015, A&A, 578, A31 Lambrechts M., Johansen A., 2014, A&A, 572, A107 Rice W. K. M., Armitage P. J., 2009, MNRAS, 396, 2228 Law N. M., et al., 2014, ApJ, 791, 35 Rice W. K. M., Veljanoski J., Collier Cameron A., 2012, MNRAS, 425, Lenzen R., et al., 2003, in Iye M., Moorwood A. F. M., eds, Proc. SPIEVol. 2567 4841, Instrument Design and Performance for Optical/Infrared Ground- Rice K., Lopez E., Forgan D., Biller B., 2015, MNRAS, 454, 1940 based Telescopes. pp 944–952, doi:10.1117/12.460044 Riddle R. L., et al., 2015, ApJ, 799, 4 Lewis N. K., et al., 2013, ApJ, 766, 95 Roberts L. C., Mason B. D., 2018, MNRAS, 473, 4497 Lidov M. L., 1962, Planet. Space Sci., 9, 719 Roberts Jr. L. C., Turner N. H., ten Brummelaar T. A., Mason B. D., Lindegren L., et al., 2018, A&A, 616, A2 Hartkopf W. I., 2011,AJ, 142, 175 Liu Y. J., et al., 2008, ApJ, 672, 553 Roberts Jr. L. C., Tokovinin A., Mason B. D., Riddle R. L., Hartkopf W. I., Lowrance P. J., Kirkpatrick J. D., Beichman C. A., 2002, ApJ, 572, L79 Law N. M., Baranec C., 2015,AJ, 149, 118 Ma B., Ge J., 2014, Monthly Notices of the Royal Astronomical Society, Roell T., Neuhäuser R., Seifahrt A., Mugrauer M., 2012, A&A, 542, A92 439, 2781 Rousset G., et al., 2003, in Wizinowich P. L., Bonaccini D., eds, Ma B., et al., 2016,AJ, 152, 112 Proc. SPIEVol. 4839, Adaptive Optical System Technologies II. pp Makarov V. V., Kaplan G. H., 2005,AJ, 129, 2420 140–149, doi:10.1117/12.459332 Maldonado J., Villaver E., 2017, A&A, 602, A38 Santerne A., et al., 2012, A&A, 544, L12 Marcy G. W., Butler R. P., 2000, Publications of the Astronomical Society Santos N. C., et al., 2002, A&A, 392, 215 of the Pacific, 112, 137 Sato B., et al., 2008, Publications of the Astronomical Society of Japan, 60, Mardling R. A., Lin D. N. C., 2002, ApJ, 573, 829 539 Masana E., Jordi C., Ribas I., 2006, A&A, 450, 735 Sato B., et al., 2010, Publications of the Astronomical Society of Japan, 62, Mason B. D., Wycoff G. L., Hartkopf W. I., Douglass G. G., Worley C. E., 1063 2001,AJ, 122, 3466 Schlaufman K. C., 2018, ApJ, 853, 37 Matsuo T., Shibai H., Ootsubo T., Tamura M., 2007, ApJ, 662, 1282 Schmidt-Kaler T., 1982, Bulletin d’Information du Centre de Donnees Stel- Mayer L., Wadsley J., Quinn T., Stadel J., 2005, MNRAS, 363, 641 laires, 23, 2 McAlister H. A., Hartkopf W. I., Hutter D. J., Franz O. G., 1987,AJ, 93, Scott N. J., Howell S. B., Horch E. P., Everett M. E., 2018, PASP, 130, 688 054502 McArthur B. E., Benedict G. F., Barnes R., Martioli E., Korzennik S., Nelan Shatsky N., 2001, A&A, 380, 238 E., Butler R. P., 2010, ApJ, 715, 1203 Shaya E. J., Olling R. P., 2011, ApJS, 192, 2 Mitchell D. S., Reffert S., Trifonov T., Quirrenbach A., Fischer D. A., 2013, Shporer A., et al., 2014, ApJ, 788, 92 A&A, 555, A87 Siverd R. J., et al., 2012, ApJ, 761, 123 Morbey C. L., Brosterhus E. B., 1974, PASP, 86, 455 Skrutskie M. F., et al., 2006,AJ, 131, 1163 Mordasini C., 2018, Planetary Population Synthesis. p. 143, Southworth J., 2012, MNRAS, 426, 1291 doi:10.1007/978-3-319-55333-7_143 Stamatellos D., 2013, in European Physical Journal Web of Conferences. p. Mordasini C., Alibert Y., Benz W., Klahr H., Henning T., 2012, A&A, 541, 08001 (arXiv:1302.3955), doi:10.1051/epjconf/20134708001 A97 Stamatellos D., Herczeg G. J., 2015, MNRAS, 449, 3432 Morton T. D., Bryson S. T., Coughlin J. L., Rowe J. F., Ravichandran G., Stamatellos D., Whitworth A. P., 2008, A&A, 480, 879 Petigura E. A., Haas M. R., Batalha N. M., 2016, ApJ, 822, 86 Szabó G. M., et al., 2011, ApJ, 736, L4 Moutou C., Vigan A., Mesa D., Desidera S., Thébault P., Zurlo A., Salter Tamuz O., et al., 2008, A&A, 480, L33 G., 2017, A&A, 602, A87 Thebault P., Haghighipour N., 2015, Planet Formation in Binaries. pp 309– Mugrauer M., Neuhäuser R., 2009, A&A, 494, 373 340, doi:10.1007/978-3-662-45052-9_13 Mugrauer M., Neuhäuser R., Mazeh T., Guenther E., Fernández M., 2004, Tokovinin A. A., Griffin R. F., Balega Y. Y., Pluzhnik E. A., Udry S., 2000, Astronomische Nachrichten, 325, 718 Astronomy Letters, 26, 116 Mugrauer M., Neuhäuser R., Mazeh T., Guenther E., Fernández M., Broeg Tokovinin A., Thomas S., Sterzik M., Udry S., 2006, A&A, 450, 681 C., 2006, Astronomische Nachrichten, 327, 321 Turner N., ten Brummelaar T., Roberts Jr. L., Gies D., Mason B., Hartkopf Mugrauer M., Neuhäuser R., Mazeh T., 2007, A&A, 469, 755 W., 2006, in The Advanced Maui Optical and Space Surveillance Tech- Naoz S., Farr W. M., Rasio F. A., 2012, ApJ, 754, L36 nologies Conference. p. E104 Nayakshin S., 2010, MNRAS, 408, L36 Udry S., Mayor M., Naef D., Pepe F., Queloz D., Santos N. C., Burnet M., Ngo H., et al., 2015, ApJ, 800, 138 2002, A&A, 390, 267 Ngo H., et al., 2016, ApJ, 827, 8 Uyama T., et al., 2017,AJ, 153, 106 Nowak G., et al., 2017,AJ, 153, 131 Vanhollebeke E., Groenewegen M. A. T., Girardi L., 2009, A&A, 498, 95 Pál A., et al., 2010, MNRAS, 401, 2665 Wagner K., Apai D., Kasper M., Robberto M., 2015, ApJ, 813, L2 Patience J., et al., 2002, ApJ, 581, 654 Wang J., Xie J.-W., Barclay T., Fischer D. A., 2014a, ApJ, 783, 4 Petrovich C., 2015, ApJ, 799, 27 Wang J., Fischer D. A., Xie J.-W., Ciardi D. R., 2014b, ApJ, 791, 111 Pichardo B., Sparke L. S., Aguilar L. A., 2005, MNRAS, 359, 521 Wang J., Fischer D. A., Horch E. P., Huang X., 2015, ApJ, 799, 229 Pinfield D. J., Jones H. R. A., Lucas P. W., Kendall T. R., Folkes S. L., White R. J., Ghez A. M., 2001, ApJ, 556, 265 Day-Jones A. C., Chappelle R. J., Steele I. A., 2006, MNRAS, 368, Wilson J. C., Kirkpatrick J. D., Gizis J. E., Skrutskie M. F., Monet D. G., 1281 Houck J. R., 2001,AJ, 122, 1989

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 27

Wilson P. A., et al., 2016, A&A, 588, A144 Winn J. N., et al., 2008, ApJ, 683, 1076 Wittenmyer R. A., Endl M., Cochran W. D., 2007, ApJ, 654, 625 174 2MASS Wittenmyer R. A., Endl M., Cochran W. D., Levison H. F., Henry G. W., Gaia DR2 2009, ApJS, 182, 97 172 Wöllert M., Brandner W., 2015, A&A, 579, A129 Wöllert M., Brandner W., Bergfors C., Henning T., 2015, A&A, 575, A23 170 Wong I., et al., 2015, ApJ, 811, 122 168 Wu Y., Murray N. W., Ramsahai J. M., 2007, ApJ, 670, 820 Youdin A. N., Goodman J., 2005, ApJ, 620, 459 166 Zacharias N., Monet D. G., Levine S. E., Urban S. E., Gaume R., Wycoﬀ

G. L., 2004, in American Astronomical Society Meeting Abstracts. Position Angle (deg) 164 p. 1418 Ziegler C., et al., 2017,AJ, 153, 66 162 Ziegler C., et al., 2018,AJ, 156, 259 160 Zucker S., Mazeh T., 2002, ApJ, 568, L113 1.7 1.8 1.9 2.0 2.1 2.2 Zucker S., Mazeh T., Santos N. C., Udry S., Mayor M., 2003, A&A, 404, Separation (arcsec) 775 Zucker S., Mazeh T., Santos N. C., Udry S., Mayor M., 2004, A&A, 426, 695 Figure A1. Common proper motion analysis of τ Gem and its companion Zuckerman B., 2014, ApJ, 791, L27 over the ∼10 year baseline between Gaia DR2 (red cross) and the astrometry from 2004 data provided by Roberts & Mason(2018) (red square). The black line shows the motion of a background object relative to τ Gem based on the Gaia DR2 parallax and proper motion of the primary, and the red dot APPENDIX A: NOTES ON INDIVIDUAL TARGETS indicates the expected position of a background object at the epoch of the 2MASS observations. The close companion is clearly found to be comoving A1 Bound systems with our target. 11 Com (HD 107383, HIP 60202) is a common proper motion binary found in the Catalog of Components of Double and Multiple positions of of τ Gem and its companion at the two epochs, clearly Stars (CCDM; Dommanget & Nys 2000). The system has a magni- demonstrating that the two objects share common proper motion tude difference ∆V = 8.0 and an angular separation of 900. 1, corre- and thus confirming that they form a physical pair. The CCDM sponding to a projected separation of 850 AU at the distance of 11 also reports another candidate to τ Gem at 5900. However this lat- Com. From the reported magnitude difference, we infer a mass of ter source is found in the Naval Observatory Merged Astrometric 0.7 M for the secondary using the BT-Settl models (Allard et al. Dataset (Nomad-I; Zacharias et al. 2004) to have a proper motion 2012) and stellar parameters given in Table2 for the primary. inconsistent with that of the primary (see Roell et al. 2012) and we 30 Ari B (HD 16232, HIP 12184) is part of a hierarchical sys- therefore discard this candidate in our survey. tem. Along with 30 Ari A (HD 16246, HIP 12189), it forms a phys- υ And (HD 9826, HIP 7513) was found by Lowrance et al. ical pair with a projected separation of 3800. 2 or 1520 AU (Shatsky (2002) to be a wide common proper motion pair on a 5500separation 2001). Both components of the F5V+F8V 30 Ari system are in turn (750 AU). The secondary stellar component υ And B has a J-band close binaries. In addition to the 9.88 MJup planet orbiting 30 Ari B magnitude of 9.39±0.03 mag and was estimated by Lowrance et al. with a period of 335.1±2.5 days (Guenther et al. 2009), Riddle et al. (2002) to have an M4.5 V spectral type and a mass of 0.2 M . The (2015) found that 30 Ari B is also orbited by another companion, primary is the host to 4 close-in planets and substellar companions 30 Ari C, with a separation of 22 AU (000. 536). Roberts et al.(2015) (Butler et al. 1997, 1999; McArthur et al. 2010; Curiel et al. 2011). subsequently demonstrated that the B-C pair is indeed comoving This binary system is also mentioned as a physical pair in Raghavan and inferred a mass of 0.5 M for the C component, classified as et al.(2006, 2010). Patience et al.(2002) and Roberts et al.(2011) an M1V dwarf (see also Kane et al. 2015). Moreover, the primary also observed the target but did not have a sufficiently wide field component 30 Ari A is itself a spectroscopic binary with a 1.1 day of view to detect the distant companion. Neither studies report any period (Morbey & Brosterhus 1974) and a total mass of 1.32 M additional, more closely-separated candidates around υ And A. (Guenther et al. 2009). AS 205 (V866 Sco, EPIC 205249328) is an extremely young τ Gem (HD 54719, HIP 34693) is reported in the CCDM and (∼0.5 Gyr) T Tauri star part of a hierarchical system in Upper sco Washington Double Star (WDS; Mason et al. 2001) catalogues to (Reboussin et al. 2015). The K5 dwarf, and brightest component have a candidate companion at a separation of 100. 9 and a magni- of the system, was found by Ghez et al.(1993) to form a common tude V = 11 mag. The system was determined to be most likely proper motion system separated by 100. 3 (corresponding to 166 AU bound in (Mitchell et al. 2013), who estimated the companion to at the distance of the system) with a low-mass spectroscopic bi- be a K0 dwarf with a mass of 0.8 M separated by 187 AU, if nary (K7+M0; Eisner et al. 2005). Prato et al.(2003) estimated a real. Roberts & Mason(2018) recently provided astrometry for this mass of a mass ratio of q∼0.2 between the A and BC components, candidate using data obtained in 2004. They found a separation of suggesting a mass of ∼0.22 M for the binary secondary. 100. 76 at a position angle of 162.5 deg. This source is also found in HAT-P-20 has a red M-dwarf companion at a separation of Gaia DR2, although it only has a 2-parameter astrometric solution 600. 86 (500 AU) fainter by ∼2 mag (WDS catalogue), which was (position only) and therefore does not have parallax and proper mo- confirmed by Bakos et al.(2011) to form a common proper motion tion measurements. From the relative positions of the primary tar- pair using Palomar sky survey archival data. The companion was get and candidate in Gaia we are able to confirm the bound nature successfully imaged in the Lucky Imaging survey by Wöllert & of this system based on the 10 year baseline between the 2004 ob- Brandner(2015) but was missed in observations from Ngo et al. servations and Gaia measurements. Figure A1 shows the relative (2015) due to the restricted field of view of their data. From the

MNRAS 000,1–30 (2019) 28 C. Fontanive et al. reported photometry and adopting the stellar parameters in Table2, 2014; Wöllert & Brandner 2015; Kraus et al. 2016). Szabó et al. we derive a mass of 0.57 M for this companion using the BT-Settl (2011) reported and confirmed Kepler-13 as a common proper mo- models at the age of the system. tion system composed of two massive A stars, also found in the HD 41004 B (HIP 28393) was identified in Santos et al. CCDM catalogue. Santerne et al.(2012) found the secondary com- (2002) and Zucker et al.(2003) as the lowest-mass component of a ponent to be a spectroscopic binary. Johnson et al.(2014) later con- 00 K1V+M2V visual binary with a 0. 54 separation, corresponding to strained the mass of Kepler-13 C to be between 0.4−0.75 M , for a 23 AU. The system has a V-magnitude difference of 3.7 mag and total mass of 1.68±0.10 M for the BC component, and 1.72±0.10 is catalogued as a physical pair in the WDS, CCDM and Tycho- M for Kepler-13 A, respectively (Shporer et al. 2014). The A-BC Hipparcos catalogues (see Roell et al. 2012). Both components are system has a projected angular separation of 100. 15, corresponding hosts to close-in substellar companions: the 0.7 M primary, HD to a physical projected separation of 610 AU (Szabó et al. 2011; 41004 A, is orbited by a giant planet at 1.33 AU with a projected Adams et al. 2012; Law et al. 2014). mass of 2.54 ± 0.74 MJup (Zucker et al. 2004), while HD 41004 NLTT 41135 was identified by Irwin et al.(2010) as a phys- 00 B (0.4 M ) hosts a brown dwarf companion at 0.017 AU with a ically associated companion to NLTT 41136 at 2. 4 separation (55 minimum mass of 18.37 ± 0.22 MJup (Zucker et al. 2003). AU). From their characterisation of the system, the authors inferred HD 87646 A (HIP 49522) is flagged as a binary in the Ty- masses of 0.16 M for NLTT 41135 and 0.21 M for NLTT 41136, cho and Hipparcos catalogues with a magnitude difference in the respectively. Hipparcos V-band of 2.66 ± 0.97 mag. Ma et al.(2016) acquired WASP-14 was found in Wöllert et al.(2015) to have a can- high-resolution images of the system and found a separation 000. 26 didate companion at 100. 4, 5.4 magnitudes fainter than the primary (∼20 AU) between the G-dwarf primary and K-dwarf secondary. in AstraLux Norte observations at the Calar Alto 2.2 m telescope. The authors derive a mass of 1.12 ± 0.09 M for the A component Ngo et al.(2015) independently identified the same candidate and and estimate a mass ratio of q ∼ 0.5 for the system. In addition were able to confirm the source to be a common proper motion to the 12.4 MJup giant planet found at 0.117 AU around HD 87646 companion to WASP-14 with a mass of 0.33 ± 0.04 M . We also A, Ma et al.(2016) also report a 57 M Jup brown dwarf candidate found in this work a distant companion to the system at 1900 AU, companion on an eccentric 1.6 AU orbit around the primary star. identified in the Gaia DR2 catalogue. We characterise WASP-14 C 00 HD 89744 (HIP 50786) is a wide binary on a 63 separation as an 0.28 M K5 dwarf (see Section 4.2.1). first identified spectroscopically by Wilson et al.(2001) and subsequently confirmed astrometrically to form a common proper mo- A2 Unconfirmed Candidate Companions tion pair by Mugrauer et al.(2004). The large angular separation of the binary corresponds to a projected separation of ∼2460 AU. Mu- 70 Vir (HD 117176, HIP 65721) was observed by Roberts et al. grauer et al.(2004) estimated a mass of 0.08 M for the secondary, (2011) using the Advance Electro-Optical System (AEOS) tele- near the hydrogen-burning limit. Raghavan et al.(2006) estimated scope. The authors report a candidate companion at a separation an L0V spectral type for the companion. HD 89744 was also ob- of 200. 86 (52 AU) around 70 Vir, which they classify as an M5 served by Roberts et al.(2011) with Adaptive Optics on the AEOS dwarf or later. With a magnitude difference of ∆I = 11.4 ± 1.2, 00 telescope, who found a faint candidate companion at 5. 62 with a we estimate a mass of 0.08 M for this candidate using the BT- magnitude difference of ∆I = 13±2 mag that is yet to be confirmed. Settl models. Pinfield et al.(2006) reported an L-dwarf candidate Other sets of observations with PUEO-KIR at CFHT by Chauvin at 4300(848 AU) based on data from the 2MASS All Sky Catalogue. et al.(2006) or the UFTI data obtained by Mugrauer et al.(2004) Common proper motion with the primary has yet to be determined do not go deep enough at that separation to retrieve this candidate. for both candidates. We did not find the latter candidate as a Gaia Given the observed magnitude difference, we infer a mass of 0.08 DR2 source, most likely too faint for Gaia. 70 Vir had also previ- M for this candidate from the BT-Settl isochrones. ously been observed by Patience et al.(2002). Observations from HD 114762 (HIP 64426) is a WDS 300. 2 (140 AU) binary pair this survey are not deep enough to retrieve the candidate found by confirmed to be comoving by Patience et al.(2002) using data from Roberts et al.(2011) and do not have a large enough field of view to Keck/NIRC and Shane/IRCAL. Bowler et al.(2009) further char- detect the wide source from Patience et al.(2002). Given the faint acterised the system, estimating an M9 spectral type and inferring infrared 2MASS magnitude of the wide source (J = 15.84 ± 0.16 a mass of 0.09 M for the companion. The companion is also re- mag), we estimate a mass of 0.07 M for the candidate adopting ported in the Adaptive Optics survey by Roberts et al.(2011). the age of the primary and the BT-Settl isochrones. HD 156846 (HIP 84856) is reported as a wide, bound binary EPIC 219388192 is solar twin in the old Ruprecht 147 star in the WDS catalogue, with a separation of 500. 1 (250 AU). Tamuz cluster (Curtis et al. 2013; Nowak et al. 2017) which was found by et al.(2008) characterised the companion to the G0 planet host as Nowak et al.(2017) to host an eccentric transiting brown dwarf an M4 dwarf of mass 0.59 M . companion. The team acquired Subaru/IRCS+AO188 images of HD 178911 B (HIP 94075) is the fainter component of a 1600. 1 the target to search for nearby companions and found two point (790 AU) physical pair found in the Tycho-Hipparcos catalogue. sources at 600and 700. 5 with contrasts of ∆H = 7.1 mag ∆H = 7.7 The primary component HD 178911 AC is itself a 4.9 AU spectro- mag, respectively. Nowak et al.(2017) estimated that the candi- scopic binary discovered by McAlister et al.(1987). The triple sys- dates, if found to be bound, would be late-type M dwarfs with tem was established to be comoving by Tokovinin et al.(2000) and masses less than ∼0.1 M . Both sources are found in the Gaia subsequently confirmed by Raghavan et al.(2006). Tokovinin et al. DR2 catalogue with separations and position angles consistent with (2000) estimated a combined mass of 1.9 M for the AC compo- those reported by Nowak et al.(2017). However, given the rela- nent, consistent with the value reported in Mugrauer et al.(2007), tively small proper motion of the target and the short time baseline while the planet host HD 178911 B has a mass of 1 M (Mugrauer between Gaia DR2 and the direct imaging data (∼1 year), we are et al. 2007; Bonfanti et al. 2016). not able to confirm or refute either of those candidates. Curtis et al. Kepler-13 A (KOI-13) has been extensively targeted with di- independently studied the same object and found 4 km s−1 offset rect imaging (Adams et al. 2012; Law et al. 2014; Shporer et al. between the center-of-mass radial velocity of the star and Ruprecht

MNRAS 000,1–30 (2019) The binarity of massive close-in giant planets 29

147’s bulk velocity (announced at the Cool Stars 19 workshop 4). DR2 epoch (2015.5) and the observations from Wöllert & Brand- As the star’s proper motion supports its cluster membership, Cur- ner(2015) (March 2015), we conclude that this is indeed the same tis et al. also obtained aperture-masking interferometry with Keck source. The Gaia DR2 source (GDR2 470650457698311296) has II and uncovered a 0.52 M companion at 82 mas (24 AU) with a full 5-parameter astrometic solution and has parallax and proper a magnitude contrast ∆K = 2.24, explaining the observed offset motion measurements highly inconsistent with those of XO-3 in (Curtis et al. private communication). Gaia DR2. We thus conclude that it is an unrelated background KELT-1 was found by Siverd et al.(2012) to have a faint object and rule out this candidate. candidate companion at 588 ± 1 mas (154 ± 8 AU) based on Keck/NIRC2 AO data. The relative brightness of the candidate was found to be ∆H = 5.90 ± 0.10 maf and ∆K0 = 5.59 ± 0.12 mag. A4 Null-detections The reported photometry suggests a mass of 0.2 M based on the BD+24 4697 (HIP 113698) was observed with Gemini North/NIRI BT-Settl models. Siverd et al.(2012) estimated an M4 −5 spectral as part of this survey. Our data did not reveal the presence of any type and concluded that the companion is physically associated to candidate companion within the field of view and detection limits the primary, with a ∼0.05% probability of being an unrelated back- of our observations. ground star based on Galatic models, in excellent agreement with CI Tau (EPIC 247584113) is a ∼2 Myr T Tauri star located in our estimates (see Table5). This target was more recently observed the Taurus star-forming region with an infrared excess in its SED by Coker et al.(2018) with the WIYN 3.5 m telescope and by and a dics resolved by Andrews & Williams(2007). It was observed Wöllert & Brandner(2015) with the Calar Alto 2.2 m telescope, by Uyama et al.(2017) with the Subaru Telescope, using the NIR although neither of these sets of observations were deep enough camera HiCIAO together with the AO188 adaptive optics system, the retrieve the candidate identified in Siverd et al.(2012). in quad PDI and standard ADI modes. The authors did not find any candidate companion within the 2000×2000field of view of their observations. This targets is also reported to be single in Kraus et al. A3 Rejected Candidates (2012) based on analyses of 2MASS images and in the HST young binary survey by White & Ghez(2001). HD 162020 (HIP 87330) had previously been observed with NACO HAT-P-2 (HD 147506, HIP 80076) was found by Lewis et al. by Eggenberger et al.(2007), who found two point sources within (2013) to have a long-term radial velocity trend, suggesting the 500from the star. The first, closer candidate was found by Eggen- presence of an outer companion in addition to its known 9 MJup berger et al.(2007) to be background, while the nature of the sec- planet on an eccentric 5 days orbit. Bonomo et al.(2017) subse- ond source was inconclusive. With new NACO data for this target, quently placed lower limits on the period and mass of this pos- we were able to refute the bound nature of this companion based on sible outer companion of >49.2 yrs (∼13 AU) and >39.5 MJup the Gaia DR2 astrometry of the primary and a decade-long base- based on radial velocity data. This is consistent with results from line between the archival and new observations (see Figure4). Our Knutson et al.(2014) who constrained the companion properties to proper motion analysis of HD 162020 and this companion is pre- M2 sin i = 8 − 200 MJup and a = 4 − 31 AU. Observations with sented in Section 3.3. NIRC2 on Keck II (Lewis et al. 2013; Ngo et al. 2015) and with HD 168443 (HIP 89844) was observed with SPHERE at VLT AstraLux Norte (Bergfors et al. 2013) did not reveal any compan- in the survey conducted in Moutou et al.(2017). Three point ∼ 00 ion but only excluded the presence of an equal-mass binary from sources are reported within 2. 5 of the primary in that paper. ∼10 AU and companions near the hydrogen-burning limit from Moutou et al.(2017) found that given the galactic latitude of the ∼50−100 AU. A companion responsible for the observed RV trend target and the crowded field of view at wider separations around could therefore still remain undetected. this object, the three identified sources are likely background con- HD 5891 (HIP 4715) was observed by Ginski et al.(2016) taminants due to the local environment of HD 168443. Using the with the Lucky Imaging instrument AstraLux at the Calar Alto 2.2 Trilegal galactic models (Vanhollebeke et al. 2009) and following m telescope and did not find any companion, achieving contrasts of the approach described in Section 4.1, we infer probabilities < 1% 4 mag at 100and 9.5 mag at 500. for any of these three sources to be physical associated to the pri- HD 33564 (HIP 25220) is listed in the CCDM as a 2500binary mary and do not consider them as bonafide candidates for the pur- although the 2 components display inconsistent proper motions and pose of our study. do not form a physical pair (Roell et al. 2012). Ginski et al.(2012) XO-3 has a faint candidate companion (i = 18.43 mag) first acquired observations of HD 33564 and excluded the presence of reported in Bergfors et al.(2013). The widely-separated candidate − 00 companions down to the substellar limit on separations of 20 100 (6 or 1500 AU projected separation) was found by Bergfors et al. AU. This star is also reported as a single object in Eggleton & (2013) to likely be a physically unrelated background object if it is a Tokovinin(2008). main-sequence star based on a colour analysis, although the authors HD 77065 (HIP 44259) is one of the two targets we observed mention the possibility of a coeval white dwarf. This target was ob- with NIRI on Gemini North. We did not find any candidates around served with Keck/NIRC2 in Ngo et al.(2015) but was not retrieved this target in our images, ruling out companions at the hydrogen- in the field of view of the images acquired for that survey. Wöllert burning limit from separations of 5 AU and substellar companions & Brandner(2015) also imaged XO-3 in a search for wide compan- with masses > 40 MJup from 70 AU. ions and detected the same source in AstraLux Norte data. A faint HD 104985 (HIP 58952) was observed with the lucky imaging source is found in Gaia DR2 (G = 18.45 mag) at the same angular camera AstraLux on the Calar Alto 2.2-m telescope by Ginski et al. separation and position angle as the detected candidate. Given the (2012). The team did not find any candidate around this target. comparable photometry and the short timespan between the Gaia HD 134113 (HIP 74033) is part of the Arcturus moving group. We observed this target with the WIYN telescope and did not find any companions within our detection limits. HD 134113 has no 4 https://doi.org/10.5281/zenodo.58758 previous direct imaging observations reported in the literature.

MNRAS 000,1–30 (2019) 30 C. Fontanive et al.

HD 156279 (HIP 84171) was observed by Ginski et al.(2016) be Gaia DR2 sources, but the fainter component only had a two- with the AstraLux instrument on the Calar Alto 2.2-m telescope. parameter astrometric solution (position only) rather than the full No companion was detected in the obtained lucky imaging data. 5-parameter solution (position, parallax and proper motion). With HD 160508 (HIP 86394) was observed as part in this work us- no parallax and proper motion measurements, we were not able to ing the WIYN imaging facilities. We did not detect any companions select these systems in our analysis and we attribute the fact that we around this object within the field of view of our images. missed them to the remaining incompleteness of Gaia DR2 and not HD 180314 (HIP 94576) was targeted by Ginski et al.(2016) to our selection criteria. We thus conclude that our selection method with lucky imaging at Calar Alto. No source was uncovered in the successfully identified all known binaries that were recoverable. obtained data within 1200, down to low-mass stellar companions. HD 203949 (HIP 105854) was observed with VLT/SPHERE in Moutou et al.(2017). That survey does not report the detection B2 Binaries with excessive astrometric disparities of any candidates around this target. Table6 reports the relative di fferences in parallax and proper mo- WASP-18 (HD 10069, HIP 7562) was part of our observed tion, together with their associated uncertainties, obtained for all sample and no source was detected in the field of view of our im- identified Gaia binaries (see also Figure7). While the majority of ages. This object had already been observed with Keck II/NIRC2 the errors are comparable in size to the calculated values them- in Ngo et al.(2015). No candidate was reported around WASP-18 selves, all systems remain fully within our arbitrary cuts at 20% at in that survey. We achieved a better contrast than that reported in the 1-σ level (with the exception of the newly discovered WASP-18 Ngo et al.(2015) both at di ffraction and background-limited sepa- AB system which is discussed in Section 4.2.2). rations and our observations allowed us to rule out the presence of A number of binaries in Table4 are part of hierarchical sys- lower-mass companions around WASP-18. A comoving object was tems and we find that 4 of the 9 previously known Gaia systems however found in this work in GDR2 at 3300 AU, outside the field have an unresolved component in Gaia DR2 (30 Ari BC, AS 205 00 of view of the direct imaging data (26. 7), for which we estimated a BC, HD 178911 AC and Kepler-13 BC), which correspond to the spectral type of M7.5 and a mass of 0.092 M (see Section 4.2.2). blue stars in Figure7. Looking at the positions of these specific systems in the parameter-space from Figure7, we find that they have the largest relative offsets in parallax and/or proper motion, and are the only systems for which the relative difference in proper APPENDIX B: GAIA DR2 ANALYSIS motion was larger than our 20% threshold in one of the coordinates In Section 4.2 we searched for sources in the Gaia DR2 catalogue (outside the shaded area in Figure7). with fractional differences of less than 20% in parallax and at least This is consistent with the idea that unresolved components one proper motion component relative to the Gaia astrometry of can have a significant effect on the measured astrometry of binary our targets. Using these selection constraints, we identified a total pairs, reinforcing the argument for loose constraints in order to en- of 11 binaries in Gaia DR2 among the targets in our sample, 9 sure that such hierarchical systems are not missed. With the excep- of which were previously known. We now examine those systems tion of AS 205 and HD 178911, all known binaries detectable by more carefully as well as the remaining systems from Table4 in Gaia would also have made a more stringent cut at ∼10% in the order to evaluate and refine our selection criteria, if needed. relative difference in parallax and in one of the proper motion components. Furthermore, the 5 known binaries that are not known to have an unresolved component (blue circles in Figure7) also make that 10% cut in both proper motion components. We thus conclude B1 Binary Completeness that most binaries should have relative discrepancies of <10% in all For completeness, we first searched for other known binaries in our astrometric parameters (π, µα∗ and µδ), while systems agreeing to sample that may have been missed by our chosen constraints. A within 20% in parallax and in one of the proper motion coordinates total of 7 known, comoving systems are missing from our identified are likely to be hierarchical systems with an unresolved component. Gaia binaries, corresponding to the companions with no parallax or We note that such wide companions are not necessarily proper motion listed in Table4. From those, 30 Ari BC, HD 41004 presently bound systems. Formerly physically associated compo- AB and HD 87646 AB have angular separations <100, the resolving nents of a binary system may continue to travel along a nearly limit of Gaia DR2, and were therefore missed because of angular identical trajectory. However, we are seeking in this study com- resolution limitations. panions that may have affected the formation or early evolution of While near-equal brightness binaries (∆G < 1 mag) are typ- inner companions and therefore also consider as bonafide any pair ically resolved with Gaia from separations of ∼100(e.g. Kepler-13, that previously constituted a bound system. We also point out that 100. 15 separation, ∆mag = 0.2 mag; AS 205 AB, 100. 3, ∆G = 1 mag), such a configuration would likely result in small discrepancies in larger separations are required to resolve lower mass ratio binaries. the observed astrometric parameters of the individual components, Ziegler et al.(2018) estimated that companions with ∆G down to an additional argument for the loose constraints considered above. ∼6 mag are consistently recovered at separations of 300−500, with In conclusion we trust that systems passing the selection criteria a roughly linear decrease in the recoverable magnitude difference described above have consistent astrometric parameters and kine- between 100−300. Based on these results, it is not surprising that sys- matics, and may be treated as binaries for the purpose of this work. tems such as WASP-14 AB (100. 45, ∆J=5.2 mag) and HD 114762 00 AB (3. 2, ∆J=7.6 mag) are not retrieved in Gaia DR2. We thus This paper has been typeset from a TEX/LATEX file prepared by the author. conclude that these companions are missing from our Gaia binary sample because they are fainter than the completeness level of Gaia DR2. Finally, the last missing binaries are τ Gem AB and HAT-P- 20 AB. In both cases, the two binary components were found to

MNRAS 000,1–30 (2019)